first commit

AIDetector_pytorch.py

@@ -0,0 +1,68 @@
+import torch
+import numpy as np
+from models.experimental import attempt_load
+from utils.general import non_max_suppression, scale_coords, letterbox
+from utils.torch_utils import select_device
+from utils.BaseDetector import baseDet
+
+
+class Detector(baseDet):
+
+    def __init__(self):
+        super(Detector, self).__init__()
+        self.init_model()
+        self.build_config()
+
+    def init_model(self):
+
+        self.weights = 'weights/yolov5m.pt'
+        self.device = '0' if torch.cuda.is_available() else 'cpu'
+        self.device = select_device(self.device)
+        model = attempt_load(self.weights, map_location=self.device)
+        model.to(self.device).eval()
+        model.half()
+        # torch.save(model, 'test.pt')
+        self.m = model
+        self.names = model.module.names if hasattr(
+            model, 'module') else model.names
+
+    def preprocess(self, img):
+
+        img0 = img.copy()
+        img = letterbox(img, new_shape=self.img_size)[0]
+        img = img[:, :, ::-1].transpose(2, 0, 1)
+        img = np.ascontiguousarray(img)
+        img = torch.from_numpy(img).to(self.device)
+        img = img.half()  # 半精度
+        img /= 255.0  # 图像归一化
+        if img.ndimension() == 3:
+            img = img.unsqueeze(0)
+
+        return img0, img
+
+    def detect(self, im):
+
+        im0, img = self.preprocess(im)
+
+        pred = self.m(img, augment=False)[0]
+        pred = pred.float()
+        pred = non_max_suppression(pred, self.threshold, 0.4)
+
+        pred_boxes = []
+        for det in pred:
+
+            if det is not None and len(det):
+                det[:, :4] = scale_coords(
+                    img.shape[2:], det[:, :4], im0.shape).round()
+
+                for *x, conf, cls_id in det:
+                    lbl = self.names[int(cls_id)]
+                    if not lbl in ['person', 'car', 'truck']:
+                        continue
+                    x1, y1 = int(x[0]), int(x[1])
+                    x2, y2 = int(x[2]), int(x[3])
+                    pred_boxes.append(
+                        (x1, y1, x2, y2, lbl, conf))
+
+        return im, pred_boxes
+

deep_sort/configs/deep_sort.yaml

@@ -0,0 +1,10 @@
+DEEPSORT:
+  REID_CKPT: "deep_sort/deep_sort/deep/checkpoint/ckpt.t7"
+  MAX_DIST: 0.2
+  MIN_CONFIDENCE: 0.3
+  NMS_MAX_OVERLAP: 0.5
+  MAX_IOU_DISTANCE: 0.7
+  MAX_AGE: 70
+  N_INIT: 3
+  NN_BUDGET: 100
+  

deep_sort/deep_sort/README.md

@@ -0,0 +1,3 @@
+# Deep Sort 
+
+This is the implemention of deep sort with pytorch.

deep_sort/deep_sort/__init__.py

@@ -0,0 +1,21 @@
+from .deep_sort import DeepSort
+
+
+__all__ = ['DeepSort', 'build_tracker']
+
+
+def build_tracker(cfg, use_cuda):
+    return DeepSort(cfg.DEEPSORT.REID_CKPT, 
+                max_dist=cfg.DEEPSORT.MAX_DIST, min_confidence=cfg.DEEPSORT.MIN_CONFIDENCE, 
+                nms_max_overlap=cfg.DEEPSORT.NMS_MAX_OVERLAP, max_iou_distance=cfg.DEEPSORT.MAX_IOU_DISTANCE, 
+                max_age=cfg.DEEPSORT.MAX_AGE, n_init=cfg.DEEPSORT.N_INIT, nn_budget=cfg.DEEPSORT.NN_BUDGET, use_cuda=use_cuda)
+    
+
+
+
+
+
+
+
+
+

deep_sort/deep_sort/deep/evaluate.py

@@ -0,0 +1,15 @@
+import torch
+
+features = torch.load("features.pth")
+qf = features["qf"]
+ql = features["ql"]
+gf = features["gf"]
+gl = features["gl"]
+
+scores = qf.mm(gf.t())
+res = scores.topk(5, dim=1)[1][:,0]
+top1correct = gl[res].eq(ql).sum().item()
+
+print("Acc top1:{:.3f}".format(top1correct/ql.size(0)))
+
+

deep_sort/deep_sort/deep/feature_extractor.py

@@ -0,0 +1,55 @@
+import torch
+import torchvision.transforms as transforms
+import numpy as np
+import cv2
+import logging
+
+from .model import Net
+
+class Extractor(object):
+    def __init__(self, model_path, use_cuda=True):
+        self.net = Net(reid=True)
+        self.device = "cuda" if torch.cuda.is_available() and use_cuda else "cpu"
+        state_dict = torch.load(model_path, map_location=lambda storage, loc: storage)['net_dict']
+        self.net.load_state_dict(state_dict)
+        logger = logging.getLogger("root.tracker")
+        logger.info("Loading weights from {}... Done!".format(model_path))
+        self.net.to(self.device)
+        self.size = (64, 128)
+        self.norm = transforms.Compose([
+            transforms.ToTensor(),
+            transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),
+        ])
+        
+
+
+    def _preprocess(self, im_crops):
+        """
+        TODO:
+            1. to float with scale from 0 to 1
+            2. resize to (64, 128) as Market1501 dataset did
+            3. concatenate to a numpy array
+            3. to torch Tensor
+            4. normalize
+        """
+        def _resize(im, size):
+            return cv2.resize(im.astype(np.float32)/255., size)
+
+        im_batch = torch.cat([self.norm(_resize(im, self.size)).unsqueeze(0) for im in im_crops], dim=0).float()
+        return im_batch
+
+
+    def __call__(self, im_crops):
+        im_batch = self._preprocess(im_crops)
+        with torch.no_grad():
+            im_batch = im_batch.to(self.device)
+            features = self.net(im_batch)
+        return features.cpu().numpy()
+
+
+if __name__ == '__main__':
+    img = cv2.imread("demo.jpg")[:,:,(2,1,0)]
+    extr = Extractor("checkpoint/ckpt.t7")
+    feature = extr(img)
+    print(feature.shape)
+

deep_sort/deep_sort/deep/model.py

@@ -0,0 +1,104 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+class BasicBlock(nn.Module):
+    def __init__(self, c_in, c_out,is_downsample=False):
+        super(BasicBlock,self).__init__()
+        self.is_downsample = is_downsample
+        if is_downsample:
+            self.conv1 = nn.Conv2d(c_in, c_out, 3, stride=2, padding=1, bias=False)
+        else:
+            self.conv1 = nn.Conv2d(c_in, c_out, 3, stride=1, padding=1, bias=False)
+        self.bn1 = nn.BatchNorm2d(c_out)
+        self.relu = nn.ReLU(True)
+        self.conv2 = nn.Conv2d(c_out,c_out,3,stride=1,padding=1, bias=False)
+        self.bn2 = nn.BatchNorm2d(c_out)
+        if is_downsample:
+            self.downsample = nn.Sequential(
+                nn.Conv2d(c_in, c_out, 1, stride=2, bias=False),
+                nn.BatchNorm2d(c_out)
+            )
+        elif c_in != c_out:
+            self.downsample = nn.Sequential(
+                nn.Conv2d(c_in, c_out, 1, stride=1, bias=False),
+                nn.BatchNorm2d(c_out)
+            )
+            self.is_downsample = True
+
+    def forward(self,x):
+        y = self.conv1(x)
+        y = self.bn1(y)
+        y = self.relu(y)
+        y = self.conv2(y)
+        y = self.bn2(y)
+        if self.is_downsample:
+            x = self.downsample(x)
+        return F.relu(x.add(y),True)
+
+def make_layers(c_in,c_out,repeat_times, is_downsample=False):
+    blocks = []
+    for i in range(repeat_times):
+        if i ==0:
+            blocks += [BasicBlock(c_in,c_out, is_downsample=is_downsample),]
+        else:
+            blocks += [BasicBlock(c_out,c_out),]
+    return nn.Sequential(*blocks)
+
+class Net(nn.Module):
+    def __init__(self, num_classes=751 ,reid=False):
+        super(Net,self).__init__()
+        # 3 128 64
+        self.conv = nn.Sequential(
+            nn.Conv2d(3,64,3,stride=1,padding=1),
+            nn.BatchNorm2d(64),
+            nn.ReLU(inplace=True),
+            # nn.Conv2d(32,32,3,stride=1,padding=1),
+            # nn.BatchNorm2d(32),
+            # nn.ReLU(inplace=True),
+            nn.MaxPool2d(3,2,padding=1),
+        )
+        # 32 64 32
+        self.layer1 = make_layers(64,64,2,False)
+        # 32 64 32
+        self.layer2 = make_layers(64,128,2,True)
+        # 64 32 16
+        self.layer3 = make_layers(128,256,2,True)
+        # 128 16 8
+        self.layer4 = make_layers(256,512,2,True)
+        # 256 8 4
+        self.avgpool = nn.AvgPool2d((8,4),1)
+        # 256 1 1 
+        self.reid = reid
+        self.classifier = nn.Sequential(
+            nn.Linear(512, 256),
+            nn.BatchNorm1d(256),
+            nn.ReLU(inplace=True),
+            nn.Dropout(),
+            nn.Linear(256, num_classes),
+        )
+    
+    def forward(self, x):
+        x = self.conv(x)
+        x = self.layer1(x)
+        x = self.layer2(x)
+        x = self.layer3(x)
+        x = self.layer4(x)
+        x = self.avgpool(x)
+        x = x.view(x.size(0),-1)
+        # B x 128
+        if self.reid:
+            x = x.div(x.norm(p=2,dim=1,keepdim=True))
+            return x
+        # classifier
+        x = self.classifier(x)
+        return x
+
+
+if __name__ == '__main__':
+    net = Net()
+    x = torch.randn(4,3,128,64)
+    y = net(x)
+    import ipdb; ipdb.set_trace()
+
+

deep_sort/deep_sort/deep/original_model.py

@@ -0,0 +1,106 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+class BasicBlock(nn.Module):
+    def __init__(self, c_in, c_out,is_downsample=False):
+        super(BasicBlock,self).__init__()
+        self.is_downsample = is_downsample
+        if is_downsample:
+            self.conv1 = nn.Conv2d(c_in, c_out, 3, stride=2, padding=1, bias=False)
+        else:
+            self.conv1 = nn.Conv2d(c_in, c_out, 3, stride=1, padding=1, bias=False)
+        self.bn1 = nn.BatchNorm2d(c_out)
+        self.relu = nn.ReLU(True)
+        self.conv2 = nn.Conv2d(c_out,c_out,3,stride=1,padding=1, bias=False)
+        self.bn2 = nn.BatchNorm2d(c_out)
+        if is_downsample:
+            self.downsample = nn.Sequential(
+                nn.Conv2d(c_in, c_out, 1, stride=2, bias=False),
+                nn.BatchNorm2d(c_out)
+            )
+        elif c_in != c_out:
+            self.downsample = nn.Sequential(
+                nn.Conv2d(c_in, c_out, 1, stride=1, bias=False),
+                nn.BatchNorm2d(c_out)
+            )
+            self.is_downsample = True
+
+    def forward(self,x):
+        y = self.conv1(x)
+        y = self.bn1(y)
+        y = self.relu(y)
+        y = self.conv2(y)
+        y = self.bn2(y)
+        if self.is_downsample:
+            x = self.downsample(x)
+        return F.relu(x.add(y),True)
+
+def make_layers(c_in,c_out,repeat_times, is_downsample=False):
+    blocks = []
+    for i in range(repeat_times):
+        if i ==0:
+            blocks += [BasicBlock(c_in,c_out, is_downsample=is_downsample),]
+        else:
+            blocks += [BasicBlock(c_out,c_out),]
+    return nn.Sequential(*blocks)
+
+class Net(nn.Module):
+    def __init__(self, num_classes=625 ,reid=False):
+        super(Net,self).__init__()
+        # 3 128 64
+        self.conv = nn.Sequential(
+            nn.Conv2d(3,32,3,stride=1,padding=1),
+            nn.BatchNorm2d(32),
+            nn.ELU(inplace=True),
+            nn.Conv2d(32,32,3,stride=1,padding=1),
+            nn.BatchNorm2d(32),
+            nn.ELU(inplace=True),
+            nn.MaxPool2d(3,2,padding=1),
+        )
+        # 32 64 32
+        self.layer1 = make_layers(32,32,2,False)
+        # 32 64 32
+        self.layer2 = make_layers(32,64,2,True)
+        # 64 32 16
+        self.layer3 = make_layers(64,128,2,True)
+        # 128 16 8
+        self.dense = nn.Sequential(
+            nn.Dropout(p=0.6),
+            nn.Linear(128*16*8, 128),
+            nn.BatchNorm1d(128),
+            nn.ELU(inplace=True)
+        )
+        # 256 1 1 
+        self.reid = reid
+        self.batch_norm = nn.BatchNorm1d(128)
+        self.classifier = nn.Sequential(
+            nn.Linear(128, num_classes),
+        )
+    
+    def forward(self, x):
+        x = self.conv(x)
+        x = self.layer1(x)
+        x = self.layer2(x)
+        x = self.layer3(x)
+
+        x = x.view(x.size(0),-1)
+        if self.reid:
+            x = self.dense[0](x)
+            x = self.dense[1](x)
+            x = x.div(x.norm(p=2,dim=1,keepdim=True))
+            return x
+        x = self.dense(x)
+        # B x 128
+        # classifier
+        x = self.classifier(x)
+        return x
+
+
+if __name__ == '__main__':
+    net = Net(reid=True)
+    x = torch.randn(4,3,128,64)
+    y = net(x)
+    import ipdb; ipdb.set_trace()
+
+

deep_sort/deep_sort/deep/test.py

@@ -0,0 +1,77 @@
+import torch
+import torch.backends.cudnn as cudnn
+import torchvision
+
+import argparse
+import os
+
+from model import Net
+
+parser = argparse.ArgumentParser(description="Train on market1501")
+parser.add_argument("--data-dir",default='data',type=str)
+parser.add_argument("--no-cuda",action="store_true")
+parser.add_argument("--gpu-id",default=0,type=int)
+args = parser.parse_args()
+
+# device
+device = "cuda:{}".format(args.gpu_id) if torch.cuda.is_available() and not args.no_cuda else "cpu"
+if torch.cuda.is_available() and not args.no_cuda:
+    cudnn.benchmark = True
+
+# data loader
+root = args.data_dir
+query_dir = os.path.join(root,"query")
+gallery_dir = os.path.join(root,"gallery")
+transform = torchvision.transforms.Compose([
+    torchvision.transforms.Resize((128,64)),
+    torchvision.transforms.ToTensor(),
+    torchvision.transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
+])
+queryloader = torch.utils.data.DataLoader(
+    torchvision.datasets.ImageFolder(query_dir, transform=transform),
+    batch_size=64, shuffle=False
+)
+galleryloader = torch.utils.data.DataLoader(
+    torchvision.datasets.ImageFolder(gallery_dir, transform=transform),
+    batch_size=64, shuffle=False
+)
+
+# net definition
+net = Net(reid=True)
+assert os.path.isfile("./checkpoint/ckpt.t7"), "Error: no checkpoint file found!"
+print('Loading from checkpoint/ckpt.t7')
+checkpoint = torch.load("./checkpoint/ckpt.t7")
+net_dict = checkpoint['net_dict']
+net.load_state_dict(net_dict, strict=False)
+net.eval()
+net.to(device)
+
+# compute features
+query_features = torch.tensor([]).float()
+query_labels = torch.tensor([]).long()
+gallery_features = torch.tensor([]).float()
+gallery_labels = torch.tensor([]).long()
+
+with torch.no_grad():
+    for idx,(inputs,labels) in enumerate(queryloader):
+        inputs = inputs.to(device)
+        features = net(inputs).cpu()
+        query_features = torch.cat((query_features, features), dim=0)
+        query_labels = torch.cat((query_labels, labels))
+
+    for idx,(inputs,labels) in enumerate(galleryloader):
+        inputs = inputs.to(device)
+        features = net(inputs).cpu()
+        gallery_features = torch.cat((gallery_features, features), dim=0)
+        gallery_labels = torch.cat((gallery_labels, labels))
+
+gallery_labels -= 2
+
+# save features
+features = {
+    "qf": query_features,
+    "ql": query_labels,
+    "gf": gallery_features,
+    "gl": gallery_labels
+}
+torch.save(features,"features.pth")

deep_sort/deep_sort/deep/train.py

@@ -0,0 +1,189 @@
+import argparse
+import os
+import time
+
+import numpy as np
+import matplotlib.pyplot as plt
+import torch
+import torch.backends.cudnn as cudnn
+import torchvision
+
+from model import Net
+
+parser = argparse.ArgumentParser(description="Train on market1501")
+parser.add_argument("--data-dir",default='data',type=str)
+parser.add_argument("--no-cuda",action="store_true")
+parser.add_argument("--gpu-id",default=0,type=int)
+parser.add_argument("--lr",default=0.1, type=float)
+parser.add_argument("--interval",'-i',default=20,type=int)
+parser.add_argument('--resume', '-r',action='store_true')
+args = parser.parse_args()
+
+# device
+device = "cuda:{}".format(args.gpu_id) if torch.cuda.is_available() and not args.no_cuda else "cpu"
+if torch.cuda.is_available() and not args.no_cuda:
+    cudnn.benchmark = True
+
+# data loading
+root = args.data_dir
+train_dir = os.path.join(root,"train")
+test_dir = os.path.join(root,"test")
+transform_train = torchvision.transforms.Compose([
+    torchvision.transforms.RandomCrop((128,64),padding=4),
+    torchvision.transforms.RandomHorizontalFlip(),
+    torchvision.transforms.ToTensor(),
+    torchvision.transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
+])
+transform_test = torchvision.transforms.Compose([
+    torchvision.transforms.Resize((128,64)),
+    torchvision.transforms.ToTensor(),
+    torchvision.transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
+])
+trainloader = torch.utils.data.DataLoader(
+    torchvision.datasets.ImageFolder(train_dir, transform=transform_train),
+    batch_size=64,shuffle=True
+)
+testloader = torch.utils.data.DataLoader(
+    torchvision.datasets.ImageFolder(test_dir, transform=transform_test),
+    batch_size=64,shuffle=True
+)
+num_classes = max(len(trainloader.dataset.classes), len(testloader.dataset.classes))
+
+# net definition
+start_epoch = 0
+net = Net(num_classes=num_classes)
+if args.resume:
+    assert os.path.isfile("./checkpoint/ckpt.t7"), "Error: no checkpoint file found!"
+    print('Loading from checkpoint/ckpt.t7')
+    checkpoint = torch.load("./checkpoint/ckpt.t7")
+    # import ipdb; ipdb.set_trace()
+    net_dict = checkpoint['net_dict']
+    net.load_state_dict(net_dict)
+    best_acc = checkpoint['acc']
+    start_epoch = checkpoint['epoch']
+net.to(device)
+
+# loss and optimizer
+criterion = torch.nn.CrossEntropyLoss()
+optimizer = torch.optim.SGD(net.parameters(), args.lr, momentum=0.9, weight_decay=5e-4)
+best_acc = 0.
+
+# train function for each epoch
+def train(epoch):
+    print("\nEpoch : %d"%(epoch+1))
+    net.train()
+    training_loss = 0.
+    train_loss = 0.
+    correct = 0
+    total = 0
+    interval = args.interval
+    start = time.time()
+    for idx, (inputs, labels) in enumerate(trainloader):
+        # forward
+        inputs,labels = inputs.to(device),labels.to(device)
+        outputs = net(inputs)
+        loss = criterion(outputs, labels)
+
+        # backward
+        optimizer.zero_grad()
+        loss.backward()
+        optimizer.step()
+
+        # accumurating
+        training_loss += loss.item()
+        train_loss += loss.item()
+        correct += outputs.max(dim=1)[1].eq(labels).sum().item()
+        total += labels.size(0)
+
+        # print 
+        if (idx+1)%interval == 0:
+            end = time.time()
+            print("[progress:{:.1f}%]time:{:.2f}s Loss:{:.5f} Correct:{}/{} Acc:{:.3f}%".format(
+                100.*(idx+1)/len(trainloader), end-start, training_loss/interval, correct, total, 100.*correct/total
+            ))
+            training_loss = 0.
+            start = time.time()
+    
+    return train_loss/len(trainloader), 1.- correct/total
+
+def test(epoch):
+    global best_acc
+    net.eval()
+    test_loss = 0.
+    correct = 0
+    total = 0
+    start = time.time()
+    with torch.no_grad():
+        for idx, (inputs, labels) in enumerate(testloader):
+            inputs, labels = inputs.to(device), labels.to(device)
+            outputs = net(inputs)
+            loss = criterion(outputs, labels)
+
+            test_loss += loss.item()
+            correct += outputs.max(dim=1)[1].eq(labels).sum().item()
+            total += labels.size(0)
+        
+        print("Testing ...")
+        end = time.time()
+        print("[progress:{:.1f}%]time:{:.2f}s Loss:{:.5f} Correct:{}/{} Acc:{:.3f}%".format(
+                100.*(idx+1)/len(testloader), end-start, test_loss/len(testloader), correct, total, 100.*correct/total
+            ))
+
+    # saving checkpoint
+    acc = 100.*correct/total
+    if acc > best_acc:
+        best_acc = acc
+        print("Saving parameters to checkpoint/ckpt.t7")
+        checkpoint = {
+            'net_dict':net.state_dict(),
+            'acc':acc,
+            'epoch':epoch,
+        }
+        if not os.path.isdir('checkpoint'):
+            os.mkdir('checkpoint')
+        torch.save(checkpoint, './checkpoint/ckpt.t7')
+
+    return test_loss/len(testloader), 1.- correct/total
+
+# plot figure
+x_epoch = []
+record = {'train_loss':[], 'train_err':[], 'test_loss':[], 'test_err':[]}
+fig = plt.figure()
+ax0 = fig.add_subplot(121, title="loss")
+ax1 = fig.add_subplot(122, title="top1err")
+def draw_curve(epoch, train_loss, train_err, test_loss, test_err):
+    global record
+    record['train_loss'].append(train_loss)
+    record['train_err'].append(train_err)
+    record['test_loss'].append(test_loss)
+    record['test_err'].append(test_err)
+
+    x_epoch.append(epoch)
+    ax0.plot(x_epoch, record['train_loss'], 'bo-', label='train')
+    ax0.plot(x_epoch, record['test_loss'], 'ro-', label='val')
+    ax1.plot(x_epoch, record['train_err'], 'bo-', label='train')
+    ax1.plot(x_epoch, record['test_err'], 'ro-', label='val')
+    if epoch == 0:
+        ax0.legend()
+        ax1.legend()
+    fig.savefig("train.jpg")
+
+# lr decay
+def lr_decay():
+    global optimizer
+    for params in optimizer.param_groups:
+        params['lr'] *= 0.1
+        lr = params['lr']
+        print("Learning rate adjusted to {}".format(lr))
+
+def main():
+    for epoch in range(start_epoch, start_epoch+40):
+        train_loss, train_err = train(epoch)
+        test_loss, test_err = test(epoch)
+        draw_curve(epoch, train_loss, train_err, test_loss, test_err)
+        if (epoch+1)%20==0:
+            lr_decay()
+
+
+if __name__ == '__main__':
+    main()

deep_sort/deep_sort/deep_sort.py

@@ -0,0 +1,115 @@
+import numpy as np
+import torch
+
+from .deep.feature_extractor import Extractor
+from .sort.nn_matching import NearestNeighborDistanceMetric
+from .sort.preprocessing import non_max_suppression
+from .sort.detection import Detection
+from .sort.tracker import Tracker
+
+
+__all__ = ['DeepSort']
+
+
+class DeepSort(object):
+    def __init__(self, model_path, max_dist=0.2, min_confidence=0.3, nms_max_overlap=1.0, max_iou_distance=0.7, max_age=70, n_init=3, nn_budget=100, use_cuda=True):
+        self.min_confidence = min_confidence
+        self.nms_max_overlap = nms_max_overlap
+
+        self.extractor = Extractor(model_path, use_cuda=use_cuda)
+
+        max_cosine_distance = max_dist
+        nn_budget = 100
+        metric = NearestNeighborDistanceMetric("cosine", max_cosine_distance, nn_budget)
+        self.tracker = Tracker(metric, max_iou_distance=max_iou_distance, max_age=max_age, n_init=n_init)
+
+    def update(self, bbox_xywh, confidences, ori_img):
+        self.height, self.width = ori_img.shape[:2]
+        # generate detections
+        features = self._get_features(bbox_xywh, ori_img)
+        bbox_tlwh = self._xywh_to_tlwh(bbox_xywh)
+        detections = [Detection(bbox_tlwh[i], conf, features[i]) for i,conf in enumerate(confidences) if conf>self.min_confidence]
+
+        # run on non-maximum supression
+        boxes = np.array([d.tlwh for d in detections])
+        scores = np.array([d.confidence for d in detections])
+        indices = non_max_suppression(boxes, self.nms_max_overlap, scores)
+        detections = [detections[i] for i in indices]
+
+        # update tracker
+        self.tracker.predict()
+        self.tracker.update(detections)
+
+        # output bbox identities
+        outputs = []
+        for track in self.tracker.tracks:
+            if not track.is_confirmed() or track.time_since_update > 1:
+                continue
+            box = track.to_tlwh()
+            x1,y1,x2,y2 = self._tlwh_to_xyxy(box)
+            track_id = track.track_id
+            outputs.append(np.array([x1,y1,x2,y2,track_id], dtype=np.int))
+        if len(outputs) > 0:
+            outputs = np.stack(outputs,axis=0)
+        return outputs
+
+
+    """
+    TODO:
+        Convert bbox from xc_yc_w_h to xtl_ytl_w_h
+    Thanks JieChen91@github.com for reporting this bug!
+    """
+    @staticmethod
+    def _xywh_to_tlwh(bbox_xywh):
+        if isinstance(bbox_xywh, np.ndarray):
+            bbox_tlwh = bbox_xywh.copy()
+        elif isinstance(bbox_xywh, torch.Tensor):
+            bbox_tlwh = bbox_xywh.clone()
+        bbox_tlwh[:,0] = bbox_xywh[:,0] - bbox_xywh[:,2]/2.
+        bbox_tlwh[:,1] = bbox_xywh[:,1] - bbox_xywh[:,3]/2.
+        return bbox_tlwh
+
+
+    def _xywh_to_xyxy(self, bbox_xywh):
+        x,y,w,h = bbox_xywh
+        x1 = max(int(x-w/2),0)
+        x2 = min(int(x+w/2),self.width-1)
+        y1 = max(int(y-h/2),0)
+        y2 = min(int(y+h/2),self.height-1)
+        return x1,y1,x2,y2
+
+    def _tlwh_to_xyxy(self, bbox_tlwh):
+        """
+        TODO:
+            Convert bbox from xtl_ytl_w_h to xc_yc_w_h
+        Thanks JieChen91@github.com for reporting this bug!
+        """
+        x,y,w,h = bbox_tlwh
+        x1 = max(int(x),0)
+        x2 = min(int(x+w),self.width-1)
+        y1 = max(int(y),0)
+        y2 = min(int(y+h),self.height-1)
+        return x1,y1,x2,y2
+
+    def _xyxy_to_tlwh(self, bbox_xyxy):
+        x1,y1,x2,y2 = bbox_xyxy
+
+        t = x1
+        l = y1
+        w = int(x2-x1)
+        h = int(y2-y1)
+        return t,l,w,h
+    
+    def _get_features(self, bbox_xywh, ori_img):
+        im_crops = []
+        for box in bbox_xywh:
+            x1,y1,x2,y2 = self._xywh_to_xyxy(box)
+            im = ori_img[y1:y2,x1:x2]
+            im_crops.append(im)
+        if im_crops:
+            features = self.extractor(im_crops)
+        else:
+            features = np.array([])
+        return features
+
+

deep_sort/deep_sort/sort/detection.py

@@ -0,0 +1,49 @@
+# vim: expandtab:ts=4:sw=4
+import numpy as np
+
+
+class Detection(object):
+    """
+    This class represents a bounding box detection in a single image.
+
+    Parameters
+    ----------
+    tlwh : array_like
+        Bounding box in format `(x, y, w, h)`.
+    confidence : float
+        Detector confidence score.
+    feature : array_like
+        A feature vector that describes the object contained in this image.
+
+    Attributes
+    ----------
+    tlwh : ndarray
+        Bounding box in format `(top left x, top left y, width, height)`.
+    confidence : ndarray
+        Detector confidence score.
+    feature : ndarray | NoneType
+        A feature vector that describes the object contained in this image.
+
+    """
+
+    def __init__(self, tlwh, confidence, feature):
+        self.tlwh = np.asarray(tlwh, dtype=np.float)
+        self.confidence = float(confidence)
+        self.feature = np.asarray(feature, dtype=np.float32)
+
+    def to_tlbr(self):
+        """Convert bounding box to format `(min x, min y, max x, max y)`, i.e.,
+        `(top left, bottom right)`.
+        """
+        ret = self.tlwh.copy()
+        ret[2:] += ret[:2]
+        return ret
+
+    def to_xyah(self):
+        """Convert bounding box to format `(center x, center y, aspect ratio,
+        height)`, where the aspect ratio is `width / height`.
+        """
+        ret = self.tlwh.copy()
+        ret[:2] += ret[2:] / 2
+        ret[2] /= ret[3]
+        return ret

deep_sort/deep_sort/sort/iou_matching.py

@@ -0,0 +1,81 @@
+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+from . import linear_assignment
+
+
+def iou(bbox, candidates):
+    """Computer intersection over union.
+
+    Parameters
+    ----------
+    bbox : ndarray
+        A bounding box in format `(top left x, top left y, width, height)`.
+    candidates : ndarray
+        A matrix of candidate bounding boxes (one per row) in the same format
+        as `bbox`.
+
+    Returns
+    -------
+    ndarray
+        The intersection over union in [0, 1] between the `bbox` and each
+        candidate. A higher score means a larger fraction of the `bbox` is
+        occluded by the candidate.
+
+    """
+    bbox_tl, bbox_br = bbox[:2], bbox[:2] + bbox[2:]
+    candidates_tl = candidates[:, :2]
+    candidates_br = candidates[:, :2] + candidates[:, 2:]
+
+    tl = np.c_[np.maximum(bbox_tl[0], candidates_tl[:, 0])[:, np.newaxis],
+               np.maximum(bbox_tl[1], candidates_tl[:, 1])[:, np.newaxis]]
+    br = np.c_[np.minimum(bbox_br[0], candidates_br[:, 0])[:, np.newaxis],
+               np.minimum(bbox_br[1], candidates_br[:, 1])[:, np.newaxis]]
+    wh = np.maximum(0., br - tl)
+
+    area_intersection = wh.prod(axis=1)
+    area_bbox = bbox[2:].prod()
+    area_candidates = candidates[:, 2:].prod(axis=1)
+    return area_intersection / (area_bbox + area_candidates - area_intersection)
+
+
+def iou_cost(tracks, detections, track_indices=None,
+             detection_indices=None):
+    """An intersection over union distance metric.
+
+    Parameters
+    ----------
+    tracks : List[deep_sort.track.Track]
+        A list of tracks.
+    detections : List[deep_sort.detection.Detection]
+        A list of detections.
+    track_indices : Optional[List[int]]
+        A list of indices to tracks that should be matched. Defaults to
+        all `tracks`.
+    detection_indices : Optional[List[int]]
+        A list of indices to detections that should be matched. Defaults
+        to all `detections`.
+
+    Returns
+    -------
+    ndarray
+        Returns a cost matrix of shape
+        len(track_indices), len(detection_indices) where entry (i, j) is
+        `1 - iou(tracks[track_indices[i]], detections[detection_indices[j]])`.
+
+    """
+    if track_indices is None:
+        track_indices = np.arange(len(tracks))
+    if detection_indices is None:
+        detection_indices = np.arange(len(detections))
+
+    cost_matrix = np.zeros((len(track_indices), len(detection_indices)))
+    for row, track_idx in enumerate(track_indices):
+        if tracks[track_idx].time_since_update > 1:
+            cost_matrix[row, :] = linear_assignment.INFTY_COST
+            continue
+
+        bbox = tracks[track_idx].to_tlwh()
+        candidates = np.asarray([detections[i].tlwh for i in detection_indices])
+        cost_matrix[row, :] = 1. - iou(bbox, candidates)
+    return cost_matrix

deep_sort/deep_sort/sort/kalman_filter.py

@@ -0,0 +1,229 @@
+# vim: expandtab:ts=4:sw=4
+import numpy as np
+import scipy.linalg
+
+
+"""
+Table for the 0.95 quantile of the chi-square distribution with N degrees of
+freedom (contains values for N=1, ..., 9). Taken from MATLAB/Octave's chi2inv
+function and used as Mahalanobis gating threshold.
+"""
+chi2inv95 = {
+    1: 3.8415,
+    2: 5.9915,
+    3: 7.8147,
+    4: 9.4877,
+    5: 11.070,
+    6: 12.592,
+    7: 14.067,
+    8: 15.507,
+    9: 16.919}
+
+
+class KalmanFilter(object):
+    """
+    A simple Kalman filter for tracking bounding boxes in image space.
+
+    The 8-dimensional state space
+
+        x, y, a, h, vx, vy, va, vh
+
+    contains the bounding box center position (x, y), aspect ratio a, height h,
+    and their respective velocities.
+
+    Object motion follows a constant velocity model. The bounding box location
+    (x, y, a, h) is taken as direct observation of the state space (linear
+    observation model).
+
+    """
+
+    def __init__(self):
+        ndim, dt = 4, 1.
+
+        # Create Kalman filter model matrices.
+        self._motion_mat = np.eye(2 * ndim, 2 * ndim)
+        for i in range(ndim):
+            self._motion_mat[i, ndim + i] = dt
+        self._update_mat = np.eye(ndim, 2 * ndim)
+
+        # Motion and observation uncertainty are chosen relative to the current
+        # state estimate. These weights control the amount of uncertainty in
+        # the model. This is a bit hacky.
+        self._std_weight_position = 1. / 20
+        self._std_weight_velocity = 1. / 160
+
+    def initiate(self, measurement):
+        """Create track from unassociated measurement.
+
+        Parameters
+        ----------
+        measurement : ndarray
+            Bounding box coordinates (x, y, a, h) with center position (x, y),
+            aspect ratio a, and height h.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the mean vector (8 dimensional) and covariance matrix (8x8
+            dimensional) of the new track. Unobserved velocities are initialized
+            to 0 mean.
+
+        """
+        mean_pos = measurement
+        mean_vel = np.zeros_like(mean_pos)
+        mean = np.r_[mean_pos, mean_vel]
+
+        std = [
+            2 * self._std_weight_position * measurement[3],
+            2 * self._std_weight_position * measurement[3],
+            1e-2,
+            2 * self._std_weight_position * measurement[3],
+            10 * self._std_weight_velocity * measurement[3],
+            10 * self._std_weight_velocity * measurement[3],
+            1e-5,
+            10 * self._std_weight_velocity * measurement[3]]
+        covariance = np.diag(np.square(std))
+        return mean, covariance
+
+    def predict(self, mean, covariance):
+        """Run Kalman filter prediction step.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The 8 dimensional mean vector of the object state at the previous
+            time step.
+        covariance : ndarray
+            The 8x8 dimensional covariance matrix of the object state at the
+            previous time step.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the mean vector and covariance matrix of the predicted
+            state. Unobserved velocities are initialized to 0 mean.
+
+        """
+        std_pos = [
+            self._std_weight_position * mean[3],
+            self._std_weight_position * mean[3],
+            1e-2,
+            self._std_weight_position * mean[3]]
+        std_vel = [
+            self._std_weight_velocity * mean[3],
+            self._std_weight_velocity * mean[3],
+            1e-5,
+            self._std_weight_velocity * mean[3]]
+        motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))
+
+        mean = np.dot(self._motion_mat, mean)
+        covariance = np.linalg.multi_dot((
+            self._motion_mat, covariance, self._motion_mat.T)) + motion_cov
+
+        return mean, covariance
+
+    def project(self, mean, covariance):
+        """Project state distribution to measurement space.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The state's mean vector (8 dimensional array).
+        covariance : ndarray
+            The state's covariance matrix (8x8 dimensional).
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the projected mean and covariance matrix of the given state
+            estimate.
+
+        """
+        std = [
+            self._std_weight_position * mean[3],
+            self._std_weight_position * mean[3],
+            1e-1,
+            self._std_weight_position * mean[3]]
+        innovation_cov = np.diag(np.square(std))
+
+        mean = np.dot(self._update_mat, mean)
+        covariance = np.linalg.multi_dot((
+            self._update_mat, covariance, self._update_mat.T))
+        return mean, covariance + innovation_cov
+
+    def update(self, mean, covariance, measurement):
+        """Run Kalman filter correction step.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The predicted state's mean vector (8 dimensional).
+        covariance : ndarray
+            The state's covariance matrix (8x8 dimensional).
+        measurement : ndarray
+            The 4 dimensional measurement vector (x, y, a, h), where (x, y)
+            is the center position, a the aspect ratio, and h the height of the
+            bounding box.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the measurement-corrected state distribution.
+
+        """
+        projected_mean, projected_cov = self.project(mean, covariance)
+
+        chol_factor, lower = scipy.linalg.cho_factor(
+            projected_cov, lower=True, check_finite=False)
+        kalman_gain = scipy.linalg.cho_solve(
+            (chol_factor, lower), np.dot(covariance, self._update_mat.T).T,
+            check_finite=False).T
+        innovation = measurement - projected_mean
+
+        new_mean = mean + np.dot(innovation, kalman_gain.T)
+        new_covariance = covariance - np.linalg.multi_dot((
+            kalman_gain, projected_cov, kalman_gain.T))
+        return new_mean, new_covariance
+
+    def gating_distance(self, mean, covariance, measurements,
+                        only_position=False):
+        """Compute gating distance between state distribution and measurements.
+
+        A suitable distance threshold can be obtained from `chi2inv95`. If
+        `only_position` is False, the chi-square distribution has 4 degrees of
+        freedom, otherwise 2.
+
+        Parameters
+        ----------
+        mean : ndarray
+            Mean vector over the state distribution (8 dimensional).
+        covariance : ndarray
+            Covariance of the state distribution (8x8 dimensional).
+        measurements : ndarray
+            An Nx4 dimensional matrix of N measurements, each in
+            format (x, y, a, h) where (x, y) is the bounding box center
+            position, a the aspect ratio, and h the height.
+        only_position : Optional[bool]
+            If True, distance computation is done with respect to the bounding
+            box center position only.
+
+        Returns
+        -------
+        ndarray
+            Returns an array of length N, where the i-th element contains the
+            squared Mahalanobis distance between (mean, covariance) and
+            `measurements[i]`.
+
+        """
+        mean, covariance = self.project(mean, covariance)
+        if only_position:
+            mean, covariance = mean[:2], covariance[:2, :2]
+            measurements = measurements[:, :2]
+
+        cholesky_factor = np.linalg.cholesky(covariance)
+        d = measurements - mean
+        z = scipy.linalg.solve_triangular(
+            cholesky_factor, d.T, lower=True, check_finite=False,
+            overwrite_b=True)
+        squared_maha = np.sum(z * z, axis=0)
+        return squared_maha

deep_sort/deep_sort/sort/linear_assignment.py

@@ -0,0 +1,192 @@
+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+# from sklearn.utils.linear_assignment_ import linear_assignment
+from scipy.optimize import linear_sum_assignment as linear_assignment
+from . import kalman_filter
+
+
+INFTY_COST = 1e+5
+
+
+def min_cost_matching(
+        distance_metric, max_distance, tracks, detections, track_indices=None,
+        detection_indices=None):
+    """Solve linear assignment problem.
+
+    Parameters
+    ----------
+    distance_metric : Callable[List[Track], List[Detection], List[int], List[int]) -> ndarray
+        The distance metric is given a list of tracks and detections as well as
+        a list of N track indices and M detection indices. The metric should
+        return the NxM dimensional cost matrix, where element (i, j) is the
+        association cost between the i-th track in the given track indices and
+        the j-th detection in the given detection_indices.
+    max_distance : float
+        Gating threshold. Associations with cost larger than this value are
+        disregarded.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : List[int]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above).
+    detection_indices : List[int]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above).
+
+    Returns
+    -------
+    (List[(int, int)], List[int], List[int])
+        Returns a tuple with the following three entries:
+        * A list of matched track and detection indices.
+        * A list of unmatched track indices.
+        * A list of unmatched detection indices.
+
+    """
+    if track_indices is None:
+        track_indices = np.arange(len(tracks))
+    if detection_indices is None:
+        detection_indices = np.arange(len(detections))
+
+    if len(detection_indices) == 0 or len(track_indices) == 0:
+        return [], track_indices, detection_indices  # Nothing to match.
+
+    cost_matrix = distance_metric(
+        tracks, detections, track_indices, detection_indices)
+    cost_matrix[cost_matrix > max_distance] = max_distance + 1e-5
+
+    row_indices, col_indices = linear_assignment(cost_matrix)
+
+    matches, unmatched_tracks, unmatched_detections = [], [], []
+    for col, detection_idx in enumerate(detection_indices):
+        if col not in col_indices:
+            unmatched_detections.append(detection_idx)
+    for row, track_idx in enumerate(track_indices):
+        if row not in row_indices:
+            unmatched_tracks.append(track_idx)
+    for row, col in zip(row_indices, col_indices):
+        track_idx = track_indices[row]
+        detection_idx = detection_indices[col]
+        if cost_matrix[row, col] > max_distance:
+            unmatched_tracks.append(track_idx)
+            unmatched_detections.append(detection_idx)
+        else:
+            matches.append((track_idx, detection_idx))
+    return matches, unmatched_tracks, unmatched_detections
+
+
+def matching_cascade(
+        distance_metric, max_distance, cascade_depth, tracks, detections,
+        track_indices=None, detection_indices=None):
+    """Run matching cascade.
+
+    Parameters
+    ----------
+    distance_metric : Callable[List[Track], List[Detection], List[int], List[int]) -> ndarray
+        The distance metric is given a list of tracks and detections as well as
+        a list of N track indices and M detection indices. The metric should
+        return the NxM dimensional cost matrix, where element (i, j) is the
+        association cost between the i-th track in the given track indices and
+        the j-th detection in the given detection indices.
+    max_distance : float
+        Gating threshold. Associations with cost larger than this value are
+        disregarded.
+    cascade_depth: int
+        The cascade depth, should be se to the maximum track age.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : Optional[List[int]]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above). Defaults to all tracks.
+    detection_indices : Optional[List[int]]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above). Defaults to all
+        detections.
+
+    Returns
+    -------
+    (List[(int, int)], List[int], List[int])
+        Returns a tuple with the following three entries:
+        * A list of matched track and detection indices.
+        * A list of unmatched track indices.
+        * A list of unmatched detection indices.
+
+    """
+    if track_indices is None:
+        track_indices = list(range(len(tracks)))
+    if detection_indices is None:
+        detection_indices = list(range(len(detections)))
+
+    unmatched_detections = detection_indices
+    matches = []
+    for level in range(cascade_depth):
+        if len(unmatched_detections) == 0:  # No detections left
+            break
+
+        track_indices_l = [
+            k for k in track_indices
+            if tracks[k].time_since_update == 1 + level
+        ]
+        if len(track_indices_l) == 0:  # Nothing to match at this level
+            continue
+
+        matches_l, _, unmatched_detections = \
+            min_cost_matching(
+                distance_metric, max_distance, tracks, detections,
+                track_indices_l, unmatched_detections)
+        matches += matches_l
+    unmatched_tracks = list(set(track_indices) - set(k for k, _ in matches))
+    return matches, unmatched_tracks, unmatched_detections
+
+
+def gate_cost_matrix(
+        kf, cost_matrix, tracks, detections, track_indices, detection_indices,
+        gated_cost=INFTY_COST, only_position=False):
+    """Invalidate infeasible entries in cost matrix based on the state
+    distributions obtained by Kalman filtering.
+
+    Parameters
+    ----------
+    kf : The Kalman filter.
+    cost_matrix : ndarray
+        The NxM dimensional cost matrix, where N is the number of track indices
+        and M is the number of detection indices, such that entry (i, j) is the
+        association cost between `tracks[track_indices[i]]` and
+        `detections[detection_indices[j]]`.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : List[int]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above).
+    detection_indices : List[int]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above).
+    gated_cost : Optional[float]
+        Entries in the cost matrix corresponding to infeasible associations are
+        set this value. Defaults to a very large value.
+    only_position : Optional[bool]
+        If True, only the x, y position of the state distribution is considered
+        during gating. Defaults to False.
+
+    Returns
+    -------
+    ndarray
+        Returns the modified cost matrix.
+
+    """
+    gating_dim = 2 if only_position else 4
+    gating_threshold = kalman_filter.chi2inv95[gating_dim]
+    measurements = np.asarray(
+        [detections[i].to_xyah() for i in detection_indices])
+    for row, track_idx in enumerate(track_indices):
+        track = tracks[track_idx]
+        gating_distance = kf.gating_distance(
+            track.mean, track.covariance, measurements, only_position)
+        cost_matrix[row, gating_distance > gating_threshold] = gated_cost
+    return cost_matrix

deep_sort/deep_sort/sort/nn_matching.py

@@ -0,0 +1,177 @@
+# vim: expandtab:ts=4:sw=4
+import numpy as np
+
+
+def _pdist(a, b):
+    """Compute pair-wise squared distance between points in `a` and `b`.
+
+    Parameters
+    ----------
+    a : array_like
+        An NxM matrix of N samples of dimensionality M.
+    b : array_like
+        An LxM matrix of L samples of dimensionality M.
+
+    Returns
+    -------
+    ndarray
+        Returns a matrix of size len(a), len(b) such that eleement (i, j)
+        contains the squared distance between `a[i]` and `b[j]`.
+
+    """
+    a, b = np.asarray(a), np.asarray(b)
+    if len(a) == 0 or len(b) == 0:
+        return np.zeros((len(a), len(b)))
+    a2, b2 = np.square(a).sum(axis=1), np.square(b).sum(axis=1)
+    r2 = -2. * np.dot(a, b.T) + a2[:, None] + b2[None, :]
+    r2 = np.clip(r2, 0., float(np.inf))
+    return r2
+
+
+def _cosine_distance(a, b, data_is_normalized=False):
+    """Compute pair-wise cosine distance between points in `a` and `b`.
+
+    Parameters
+    ----------
+    a : array_like
+        An NxM matrix of N samples of dimensionality M.
+    b : array_like
+        An LxM matrix of L samples of dimensionality M.
+    data_is_normalized : Optional[bool]
+        If True, assumes rows in a and b are unit length vectors.
+        Otherwise, a and b are explicitly normalized to lenght 1.
+
+    Returns
+    -------
+    ndarray
+        Returns a matrix of size len(a), len(b) such that eleement (i, j)
+        contains the squared distance between `a[i]` and `b[j]`.
+
+    """
+    if not data_is_normalized:
+        a = np.asarray(a) / np.linalg.norm(a, axis=1, keepdims=True)
+        b = np.asarray(b) / np.linalg.norm(b, axis=1, keepdims=True)
+    return 1. - np.dot(a, b.T)
+
+
+def _nn_euclidean_distance(x, y):
+    """ Helper function for nearest neighbor distance metric (Euclidean).
+
+    Parameters
+    ----------
+    x : ndarray
+        A matrix of N row-vectors (sample points).
+    y : ndarray
+        A matrix of M row-vectors (query points).
+
+    Returns
+    -------
+    ndarray
+        A vector of length M that contains for each entry in `y` the
+        smallest Euclidean distance to a sample in `x`.
+
+    """
+    distances = _pdist(x, y)
+    return np.maximum(0.0, distances.min(axis=0))
+
+
+def _nn_cosine_distance(x, y):
+    """ Helper function for nearest neighbor distance metric (cosine).
+
+    Parameters
+    ----------
+    x : ndarray
+        A matrix of N row-vectors (sample points).
+    y : ndarray
+        A matrix of M row-vectors (query points).
+
+    Returns
+    -------
+    ndarray
+        A vector of length M that contains for each entry in `y` the
+        smallest cosine distance to a sample in `x`.
+
+    """
+    distances = _cosine_distance(x, y)
+    return distances.min(axis=0)
+
+
+class NearestNeighborDistanceMetric(object):
+    """
+    A nearest neighbor distance metric that, for each target, returns
+    the closest distance to any sample that has been observed so far.
+
+    Parameters
+    ----------
+    metric : str
+        Either "euclidean" or "cosine".
+    matching_threshold: float
+        The matching threshold. Samples with larger distance are considered an
+        invalid match.
+    budget : Optional[int]
+        If not None, fix samples per class to at most this number. Removes
+        the oldest samples when the budget is reached.
+
+    Attributes
+    ----------
+    samples : Dict[int -> List[ndarray]]
+        A dictionary that maps from target identities to the list of samples
+        that have been observed so far.
+
+    """
+
+    def __init__(self, metric, matching_threshold, budget=None):
+
+
+        if metric == "euclidean":
+            self._metric = _nn_euclidean_distance
+        elif metric == "cosine":
+            self._metric = _nn_cosine_distance
+        else:
+            raise ValueError(
+                "Invalid metric; must be either 'euclidean' or 'cosine'")
+        self.matching_threshold = matching_threshold
+        self.budget = budget
+        self.samples = {}
+
+    def partial_fit(self, features, targets, active_targets):
+        """Update the distance metric with new data.
+
+        Parameters
+        ----------
+        features : ndarray
+            An NxM matrix of N features of dimensionality M.
+        targets : ndarray
+            An integer array of associated target identities.
+        active_targets : List[int]
+            A list of targets that are currently present in the scene.
+
+        """
+        for feature, target in zip(features, targets):
+            self.samples.setdefault(target, []).append(feature)
+            if self.budget is not None:
+                self.samples[target] = self.samples[target][-self.budget:]
+        self.samples = {k: self.samples[k] for k in active_targets}
+
+    def distance(self, features, targets):
+        """Compute distance between features and targets.
+
+        Parameters
+        ----------
+        features : ndarray
+            An NxM matrix of N features of dimensionality M.
+        targets : List[int]
+            A list of targets to match the given `features` against.
+
+        Returns
+        -------
+        ndarray
+            Returns a cost matrix of shape len(targets), len(features), where
+            element (i, j) contains the closest squared distance between
+            `targets[i]` and `features[j]`.
+
+        """
+        cost_matrix = np.zeros((len(targets), len(features)))
+        for i, target in enumerate(targets):
+            cost_matrix[i, :] = self._metric(self.samples[target], features)
+        return cost_matrix

deep_sort/deep_sort/sort/preprocessing.py

@@ -0,0 +1,73 @@
+# vim: expandtab:ts=4:sw=4
+import numpy as np
+import cv2
+
+
+def non_max_suppression(boxes, max_bbox_overlap, scores=None):
+    """Suppress overlapping detections.
+
+    Original code from [1]_ has been adapted to include confidence score.
+
+    .. [1] http://www.pyimagesearch.com/2015/02/16/
+           faster-non-maximum-suppression-python/
+
+    Examples
+    --------
+
+        >>> boxes = [d.roi for d in detections]
+        >>> scores = [d.confidence for d in detections]
+        >>> indices = non_max_suppression(boxes, max_bbox_overlap, scores)
+        >>> detections = [detections[i] for i in indices]
+
+    Parameters
+    ----------
+    boxes : ndarray
+        Array of ROIs (x, y, width, height).
+    max_bbox_overlap : float
+        ROIs that overlap more than this values are suppressed.
+    scores : Optional[array_like]
+        Detector confidence score.
+
+    Returns
+    -------
+    List[int]
+        Returns indices of detections that have survived non-maxima suppression.
+
+    """
+    if len(boxes) == 0:
+        return []
+
+    boxes = boxes.astype(np.float)
+    pick = []
+
+    x1 = boxes[:, 0]
+    y1 = boxes[:, 1]
+    x2 = boxes[:, 2] + boxes[:, 0]
+    y2 = boxes[:, 3] + boxes[:, 1]
+
+    area = (x2 - x1 + 1) * (y2 - y1 + 1)
+    if scores is not None:
+        idxs = np.argsort(scores)
+    else:
+        idxs = np.argsort(y2)
+
+    while len(idxs) > 0:
+        last = len(idxs) - 1
+        i = idxs[last]
+        pick.append(i)
+
+        xx1 = np.maximum(x1[i], x1[idxs[:last]])
+        yy1 = np.maximum(y1[i], y1[idxs[:last]])
+        xx2 = np.minimum(x2[i], x2[idxs[:last]])
+        yy2 = np.minimum(y2[i], y2[idxs[:last]])
+
+        w = np.maximum(0, xx2 - xx1 + 1)
+        h = np.maximum(0, yy2 - yy1 + 1)
+
+        overlap = (w * h) / area[idxs[:last]]
+
+        idxs = np.delete(
+            idxs, np.concatenate(
+                ([last], np.where(overlap > max_bbox_overlap)[0])))
+
+    return pick

deep_sort/deep_sort/sort/track.py

@@ -0,0 +1,166 @@
+# vim: expandtab:ts=4:sw=4
+
+
+class TrackState:
+    """
+    Enumeration type for the single target track state. Newly created tracks are
+    classified as `tentative` until enough evidence has been collected. Then,
+    the track state is changed to `confirmed`. Tracks that are no longer alive
+    are classified as `deleted` to mark them for removal from the set of active
+    tracks.
+
+    """
+
+    Tentative = 1
+    Confirmed = 2
+    Deleted = 3
+
+
+class Track:
+    """
+    A single target track with state space `(x, y, a, h)` and associated
+    velocities, where `(x, y)` is the center of the bounding box, `a` is the
+    aspect ratio and `h` is the height.
+
+    Parameters
+    ----------
+    mean : ndarray
+        Mean vector of the initial state distribution.
+    covariance : ndarray
+        Covariance matrix of the initial state distribution.
+    track_id : int
+        A unique track identifier.
+    n_init : int
+        Number of consecutive detections before the track is confirmed. The
+        track state is set to `Deleted` if a miss occurs within the first
+        `n_init` frames.
+    max_age : int
+        The maximum number of consecutive misses before the track state is
+        set to `Deleted`.
+    feature : Optional[ndarray]
+        Feature vector of the detection this track originates from. If not None,
+        this feature is added to the `features` cache.
+
+    Attributes
+    ----------
+    mean : ndarray
+        Mean vector of the initial state distribution.
+    covariance : ndarray
+        Covariance matrix of the initial state distribution.
+    track_id : int
+        A unique track identifier.
+    hits : int
+        Total number of measurement updates.
+    age : int
+        Total number of frames since first occurance.
+    time_since_update : int
+        Total number of frames since last measurement update.
+    state : TrackState
+        The current track state.
+    features : List[ndarray]
+        A cache of features. On each measurement update, the associated feature
+        vector is added to this list.
+
+    """
+
+    def __init__(self, mean, covariance, track_id, n_init, max_age,
+                 feature=None):
+        self.mean = mean
+        self.covariance = covariance
+        self.track_id = track_id
+        self.hits = 1
+        self.age = 1
+        self.time_since_update = 0
+
+        self.state = TrackState.Tentative
+        self.features = []
+        if feature is not None:
+            self.features.append(feature)
+
+        self._n_init = n_init
+        self._max_age = max_age
+
+    def to_tlwh(self):
+        """Get current position in bounding box format `(top left x, top left y,
+        width, height)`.
+
+        Returns
+        -------
+        ndarray
+            The bounding box.
+
+        """
+        ret = self.mean[:4].copy()
+        ret[2] *= ret[3]
+        ret[:2] -= ret[2:] / 2
+        return ret
+
+    def to_tlbr(self):
+        """Get current position in bounding box format `(min x, miny, max x,
+        max y)`.
+
+        Returns
+        -------
+        ndarray
+            The bounding box.
+
+        """
+        ret = self.to_tlwh()
+        ret[2:] = ret[:2] + ret[2:]
+        return ret
+
+    def predict(self, kf):
+        """Propagate the state distribution to the current time step using a
+        Kalman filter prediction step.
+
+        Parameters
+        ----------
+        kf : kalman_filter.KalmanFilter
+            The Kalman filter.
+
+        """
+        self.mean, self.covariance = kf.predict(self.mean, self.covariance)
+        self.age += 1
+        self.time_since_update += 1
+
+    def update(self, kf, detection):
+        """Perform Kalman filter measurement update step and update the feature
+        cache.
+
+        Parameters
+        ----------
+        kf : kalman_filter.KalmanFilter
+            The Kalman filter.
+        detection : Detection
+            The associated detection.
+
+        """
+        self.mean, self.covariance = kf.update(
+            self.mean, self.covariance, detection.to_xyah())
+        self.features.append(detection.feature)
+
+        self.hits += 1
+        self.time_since_update = 0
+        if self.state == TrackState.Tentative and self.hits >= self._n_init:
+            self.state = TrackState.Confirmed
+
+    def mark_missed(self):
+        """Mark this track as missed (no association at the current time step).
+        """
+        if self.state == TrackState.Tentative:
+            self.state = TrackState.Deleted
+        elif self.time_since_update > self._max_age:
+            self.state = TrackState.Deleted
+
+    def is_tentative(self):
+        """Returns True if this track is tentative (unconfirmed).
+        """
+        return self.state == TrackState.Tentative
+
+    def is_confirmed(self):
+        """Returns True if this track is confirmed."""
+        return self.state == TrackState.Confirmed
+
+    def is_deleted(self):
+        """Returns True if this track is dead and should be deleted."""
+        return self.state == TrackState.Deleted

deep_sort/deep_sort/sort/tracker.py

@@ -0,0 +1,138 @@
+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+from . import kalman_filter
+from . import linear_assignment
+from . import iou_matching
+from .track import Track
+
+
+class Tracker:
+    """
+    This is the multi-target tracker.
+
+    Parameters
+    ----------
+    metric : nn_matching.NearestNeighborDistanceMetric
+        A distance metric for measurement-to-track association.
+    max_age : int
+        Maximum number of missed misses before a track is deleted.
+    n_init : int
+        Number of consecutive detections before the track is confirmed. The
+        track state is set to `Deleted` if a miss occurs within the first
+        `n_init` frames.
+
+    Attributes
+    ----------
+    metric : nn_matching.NearestNeighborDistanceMetric
+        The distance metric used for measurement to track association.
+    max_age : int
+        Maximum number of missed misses before a track is deleted.
+    n_init : int
+        Number of frames that a track remains in initialization phase.
+    kf : kalman_filter.KalmanFilter
+        A Kalman filter to filter target trajectories in image space.
+    tracks : List[Track]
+        The list of active tracks at the current time step.
+
+    """
+
+    def __init__(self, metric, max_iou_distance=0.7, max_age=70, n_init=3):
+        self.metric = metric
+        self.max_iou_distance = max_iou_distance
+        self.max_age = max_age
+        self.n_init = n_init
+
+        self.kf = kalman_filter.KalmanFilter()
+        self.tracks = []
+        self._next_id = 1
+
+    def predict(self):
+        """Propagate track state distributions one time step forward.
+
+        This function should be called once every time step, before `update`.
+        """
+        for track in self.tracks:
+            track.predict(self.kf)
+
+    def update(self, detections):
+        """Perform measurement update and track management.
+
+        Parameters
+        ----------
+        detections : List[deep_sort.detection.Detection]
+            A list of detections at the current time step.
+
+        """
+        # Run matching cascade.
+        matches, unmatched_tracks, unmatched_detections = \
+            self._match(detections)
+
+        # Update track set.
+        for track_idx, detection_idx in matches:
+            self.tracks[track_idx].update(
+                self.kf, detections[detection_idx])
+        for track_idx in unmatched_tracks:
+            self.tracks[track_idx].mark_missed()
+        for detection_idx in unmatched_detections:
+            self._initiate_track(detections[detection_idx])
+        self.tracks = [t for t in self.tracks if not t.is_deleted()]
+
+        # Update distance metric.
+        active_targets = [t.track_id for t in self.tracks if t.is_confirmed()]
+        features, targets = [], []
+        for track in self.tracks:
+            if not track.is_confirmed():
+                continue
+            features += track.features
+            targets += [track.track_id for _ in track.features]
+            track.features = []
+        self.metric.partial_fit(
+            np.asarray(features), np.asarray(targets), active_targets)
+
+    def _match(self, detections):
+
+        def gated_metric(tracks, dets, track_indices, detection_indices):
+            features = np.array([dets[i].feature for i in detection_indices])
+            targets = np.array([tracks[i].track_id for i in track_indices])
+            cost_matrix = self.metric.distance(features, targets)
+            cost_matrix = linear_assignment.gate_cost_matrix(
+                self.kf, cost_matrix, tracks, dets, track_indices,
+                detection_indices)
+
+            return cost_matrix
+
+        # Split track set into confirmed and unconfirmed tracks.
+        confirmed_tracks = [
+            i for i, t in enumerate(self.tracks) if t.is_confirmed()]
+        unconfirmed_tracks = [
+            i for i, t in enumerate(self.tracks) if not t.is_confirmed()]
+
+        # Associate confirmed tracks using appearance features.
+        matches_a, unmatched_tracks_a, unmatched_detections = \
+            linear_assignment.matching_cascade(
+                gated_metric, self.metric.matching_threshold, self.max_age,
+                self.tracks, detections, confirmed_tracks)
+
+        # Associate remaining tracks together with unconfirmed tracks using IOU.
+        iou_track_candidates = unconfirmed_tracks + [
+            k for k in unmatched_tracks_a if
+            self.tracks[k].time_since_update == 1]
+        unmatched_tracks_a = [
+            k for k in unmatched_tracks_a if
+            self.tracks[k].time_since_update != 1]
+        matches_b, unmatched_tracks_b, unmatched_detections = \
+            linear_assignment.min_cost_matching(
+                iou_matching.iou_cost, self.max_iou_distance, self.tracks,
+                detections, iou_track_candidates, unmatched_detections)
+
+        matches = matches_a + matches_b
+        unmatched_tracks = list(set(unmatched_tracks_a + unmatched_tracks_b))
+        return matches, unmatched_tracks, unmatched_detections
+
+    def _initiate_track(self, detection):
+        mean, covariance = self.kf.initiate(detection.to_xyah())
+        self.tracks.append(Track(
+            mean, covariance, self._next_id, self.n_init, self.max_age,
+            detection.feature))
+        self._next_id += 1

deep_sort/utils/asserts.py

@@ -0,0 +1,13 @@
+from os import environ
+
+
+def assert_in(file, files_to_check):
+    if file not in files_to_check:
+        raise AssertionError("{} does not exist in the list".format(str(file)))
+    return True
+
+
+def assert_in_env(check_list: list):
+    for item in check_list:
+        assert_in(item, environ.keys())
+    return True

deep_sort/utils/draw.py

@@ -0,0 +1,36 @@
+import numpy as np
+import cv2
+
+palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1)
+
+
+def compute_color_for_labels(label):
+    """
+    Simple function that adds fixed color depending on the class
+    """
+    color = [int((p * (label ** 2 - label + 1)) % 255) for p in palette]
+    return tuple(color)
+
+
+def draw_boxes(img, bbox, identities=None, offset=(0,0)):
+    for i,box in enumerate(bbox):
+        x1,y1,x2,y2 = [int(i) for i in box]
+        x1 += offset[0]
+        x2 += offset[0]
+        y1 += offset[1]
+        y2 += offset[1]
+        # box text and bar
+        id = int(identities[i]) if identities is not None else 0    
+        color = compute_color_for_labels(id)
+        label = '{}{:d}'.format("", id)
+        t_size = cv2.getTextSize(label, cv2.FONT_HERSHEY_PLAIN, 2 , 2)[0]
+        cv2.rectangle(img,(x1, y1),(x2,y2),color,3)
+        cv2.rectangle(img,(x1, y1),(x1+t_size[0]+3,y1+t_size[1]+4), color,-1)
+        cv2.putText(img,label,(x1,y1+t_size[1]+4), cv2.FONT_HERSHEY_PLAIN, 2, [255,255,255], 2)
+    return img
+
+
+
+if __name__ == '__main__':
+    for i in range(82):
+        print(compute_color_for_labels(i))

deep_sort/utils/evaluation.py

@@ -0,0 +1,103 @@
+import os
+import numpy as np
+import copy
+import motmetrics as mm
+mm.lap.default_solver = 'lap'
+from utils.io import read_results, unzip_objs
+
+
+class Evaluator(object):
+
+    def __init__(self, data_root, seq_name, data_type):
+        self.data_root = data_root
+        self.seq_name = seq_name
+        self.data_type = data_type
+
+        self.load_annotations()
+        self.reset_accumulator()
+
+    def load_annotations(self):
+        assert self.data_type == 'mot'
+
+        gt_filename = os.path.join(self.data_root, self.seq_name, 'gt', 'gt.txt')
+        self.gt_frame_dict = read_results(gt_filename, self.data_type, is_gt=True)
+        self.gt_ignore_frame_dict = read_results(gt_filename, self.data_type, is_ignore=True)
+
+    def reset_accumulator(self):
+        self.acc = mm.MOTAccumulator(auto_id=True)
+
+    def eval_frame(self, frame_id, trk_tlwhs, trk_ids, rtn_events=False):
+        # results
+        trk_tlwhs = np.copy(trk_tlwhs)
+        trk_ids = np.copy(trk_ids)
+
+        # gts
+        gt_objs = self.gt_frame_dict.get(frame_id, [])
+        gt_tlwhs, gt_ids = unzip_objs(gt_objs)[:2]
+
+        # ignore boxes
+        ignore_objs = self.gt_ignore_frame_dict.get(frame_id, [])
+        ignore_tlwhs = unzip_objs(ignore_objs)[0]
+
+
+        # remove ignored results
+        keep = np.ones(len(trk_tlwhs), dtype=bool)
+        iou_distance = mm.distances.iou_matrix(ignore_tlwhs, trk_tlwhs, max_iou=0.5)
+        if len(iou_distance) > 0:
+            match_is, match_js = mm.lap.linear_sum_assignment(iou_distance)
+            match_is, match_js = map(lambda a: np.asarray(a, dtype=int), [match_is, match_js])
+            match_ious = iou_distance[match_is, match_js]
+
+            match_js = np.asarray(match_js, dtype=int)
+            match_js = match_js[np.logical_not(np.isnan(match_ious))]
+            keep[match_js] = False
+            trk_tlwhs = trk_tlwhs[keep]
+            trk_ids = trk_ids[keep]
+
+        # get distance matrix
+        iou_distance = mm.distances.iou_matrix(gt_tlwhs, trk_tlwhs, max_iou=0.5)
+
+        # acc
+        self.acc.update(gt_ids, trk_ids, iou_distance)
+
+        if rtn_events and iou_distance.size > 0 and hasattr(self.acc, 'last_mot_events'):
+            events = self.acc.last_mot_events  # only supported by https://github.com/longcw/py-motmetrics
+        else:
+            events = None
+        return events
+
+    def eval_file(self, filename):
+        self.reset_accumulator()
+
+        result_frame_dict = read_results(filename, self.data_type, is_gt=False)
+        frames = sorted(list(set(self.gt_frame_dict.keys()) | set(result_frame_dict.keys())))
+        for frame_id in frames:
+            trk_objs = result_frame_dict.get(frame_id, [])
+            trk_tlwhs, trk_ids = unzip_objs(trk_objs)[:2]
+            self.eval_frame(frame_id, trk_tlwhs, trk_ids, rtn_events=False)
+
+        return self.acc
+
+    @staticmethod
+    def get_summary(accs, names, metrics=('mota', 'num_switches', 'idp', 'idr', 'idf1', 'precision', 'recall')):
+        names = copy.deepcopy(names)
+        if metrics is None:
+            metrics = mm.metrics.motchallenge_metrics
+        metrics = copy.deepcopy(metrics)
+
+        mh = mm.metrics.create()
+        summary = mh.compute_many(
+            accs,
+            metrics=metrics,
+            names=names,
+            generate_overall=True
+        )
+
+        return summary
+
+    @staticmethod
+    def save_summary(summary, filename):
+        import pandas as pd
+        writer = pd.ExcelWriter(filename)
+        summary.to_excel(writer)
+        writer.save()

deep_sort/utils/io.py

@@ -0,0 +1,133 @@
+import os
+from typing import Dict
+import numpy as np
+
+# from utils.log import get_logger
+
+
+def write_results(filename, results, data_type):
+    if data_type == 'mot':
+        save_format = '{frame},{id},{x1},{y1},{w},{h},-1,-1,-1,-1\n'
+    elif data_type == 'kitti':
+        save_format = '{frame} {id} pedestrian 0 0 -10 {x1} {y1} {x2} {y2} -10 -10 -10 -1000 -1000 -1000 -10\n'
+    else:
+        raise ValueError(data_type)
+
+    with open(filename, 'w') as f:
+        for frame_id, tlwhs, track_ids in results:
+            if data_type == 'kitti':
+                frame_id -= 1
+            for tlwh, track_id in zip(tlwhs, track_ids):
+                if track_id < 0:
+                    continue
+                x1, y1, w, h = tlwh
+                x2, y2 = x1 + w, y1 + h
+                line = save_format.format(frame=frame_id, id=track_id, x1=x1, y1=y1, x2=x2, y2=y2, w=w, h=h)
+                f.write(line)
+
+
+# def write_results(filename, results_dict: Dict, data_type: str):
+#     if not filename:
+#         return
+#     path = os.path.dirname(filename)
+#     if not os.path.exists(path):
+#         os.makedirs(path)
+
+#     if data_type in ('mot', 'mcmot', 'lab'):
+#         save_format = '{frame},{id},{x1},{y1},{w},{h},1,-1,-1,-1\n'
+#     elif data_type == 'kitti':
+#         save_format = '{frame} {id} pedestrian -1 -1 -10 {x1} {y1} {x2} {y2} -1 -1 -1 -1000 -1000 -1000 -10 {score}\n'
+#     else:
+#         raise ValueError(data_type)
+
+#     with open(filename, 'w') as f:
+#         for frame_id, frame_data in results_dict.items():
+#             if data_type == 'kitti':
+#                 frame_id -= 1
+#             for tlwh, track_id in frame_data:
+#                 if track_id < 0:
+#                     continue
+#                 x1, y1, w, h = tlwh
+#                 x2, y2 = x1 + w, y1 + h
+#                 line = save_format.format(frame=frame_id, id=track_id, x1=x1, y1=y1, x2=x2, y2=y2, w=w, h=h, score=1.0)
+#                 f.write(line)
+#     logger.info('Save results to {}'.format(filename))
+
+
+def read_results(filename, data_type: str, is_gt=False, is_ignore=False):
+    if data_type in ('mot', 'lab'):
+        read_fun = read_mot_results
+    else:
+        raise ValueError('Unknown data type: {}'.format(data_type))
+
+    return read_fun(filename, is_gt, is_ignore)
+
+
+"""
+labels={'ped', ...			% 1
+'person_on_vhcl', ...	% 2
+'car', ...				% 3
+'bicycle', ...			% 4
+'mbike', ...			% 5
+'non_mot_vhcl', ...		% 6
+'static_person', ...	% 7
+'distractor', ...		% 8
+'occluder', ...			% 9
+'occluder_on_grnd', ...		%10
+'occluder_full', ...		% 11
+'reflection', ...		% 12
+'crowd' ...			% 13
+};
+"""
+
+
+def read_mot_results(filename, is_gt, is_ignore):
+    valid_labels = {1}
+    ignore_labels = {2, 7, 8, 12}
+    results_dict = dict()
+    if os.path.isfile(filename):
+        with open(filename, 'r') as f:
+            for line in f.readlines():
+                linelist = line.split(',')
+                if len(linelist) < 7:
+                    continue
+                fid = int(linelist[0])
+                if fid < 1:
+                    continue
+                results_dict.setdefault(fid, list())
+
+                if is_gt:
+                    if 'MOT16-' in filename or 'MOT17-' in filename:
+                        label = int(float(linelist[7]))
+                        mark = int(float(linelist[6]))
+                        if mark == 0 or label not in valid_labels:
+                            continue
+                    score = 1
+                elif is_ignore:
+                    if 'MOT16-' in filename or 'MOT17-' in filename:
+                        label = int(float(linelist[7]))
+                        vis_ratio = float(linelist[8])
+                        if label not in ignore_labels and vis_ratio >= 0:
+                            continue
+                    else:
+                        continue
+                    score = 1
+                else:
+                    score = float(linelist[6])
+
+                tlwh = tuple(map(float, linelist[2:6]))
+                target_id = int(linelist[1])
+
+                results_dict[fid].append((tlwh, target_id, score))
+
+    return results_dict
+
+
+def unzip_objs(objs):
+    if len(objs) > 0:
+        tlwhs, ids, scores = zip(*objs)
+    else:
+        tlwhs, ids, scores = [], [], []
+    tlwhs = np.asarray(tlwhs, dtype=float).reshape(-1, 4)
+
+    return tlwhs, ids, scores

deep_sort/utils/json_logger.py

@@ -0,0 +1,383 @@
+"""
+References:
+    https://medium.com/analytics-vidhya/creating-a-custom-logging-mechanism-for-real-time-object-detection-using-tdd-4ca2cfcd0a2f
+"""
+import json
+from os import makedirs
+from os.path import exists, join
+from datetime import datetime
+
+
+class JsonMeta(object):
+    HOURS = 3
+    MINUTES = 59
+    SECONDS = 59
+    PATH_TO_SAVE = 'LOGS'
+    DEFAULT_FILE_NAME = 'remaining'
+
+
+class BaseJsonLogger(object):
+    """
+    This is the base class that returns __dict__ of its own
+    it also returns the dicts of objects in the attributes that are list instances
+
+    """
+
+    def dic(self):
+        # returns dicts of objects
+        out = {}
+        for k, v in self.__dict__.items():
+            if hasattr(v, 'dic'):
+                out[k] = v.dic()
+            elif isinstance(v, list):
+                out[k] = self.list(v)
+            else:
+                out[k] = v
+        return out
+
+    @staticmethod
+    def list(values):
+        # applies the dic method on items in the list
+        return [v.dic() if hasattr(v, 'dic') else v for v in values]
+
+
+class Label(BaseJsonLogger):
+    """
+    For each bounding box there are various categories with confidences. Label class keeps track of that information.
+    """
+
+    def __init__(self, category: str, confidence: float):
+        self.category = category
+        self.confidence = confidence
+
+
+class Bbox(BaseJsonLogger):
+    """
+    This module stores the information for each frame and use them in JsonParser
+    Attributes:
+        labels (list): List of label module.
+        top (int):
+        left (int):
+        width (int):
+        height (int):
+
+    Args:
+        bbox_id (float):
+        top (int):
+        left (int):
+        width (int):
+        height (int):
+
+    References:
+        Check Label module for better understanding.
+
+
+    """
+
+    def __init__(self, bbox_id, top, left, width, height):
+        self.labels = []
+        self.bbox_id = bbox_id
+        self.top = top
+        self.left = left
+        self.width = width
+        self.height = height
+
+    def add_label(self, category, confidence):
+        # adds category and confidence only if top_k is not exceeded.
+        self.labels.append(Label(category, confidence))
+
+    def labels_full(self, value):
+        return len(self.labels) == value
+
+
+class Frame(BaseJsonLogger):
+    """
+    This module stores the information for each frame and use them in JsonParser
+    Attributes:
+        timestamp (float): The elapsed time of captured frame
+        frame_id (int): The frame number of the captured video
+        bboxes (list of Bbox objects): Stores the list of bbox objects.
+
+    References:
+        Check Bbox class for better information
+
+    Args:
+        timestamp (float):
+        frame_id (int):
+
+    """
+
+    def __init__(self, frame_id: int, timestamp: float = None):
+        self.frame_id = frame_id
+        self.timestamp = timestamp
+        self.bboxes = []
+
+    def add_bbox(self, bbox_id: int, top: int, left: int, width: int, height: int):
+        bboxes_ids = [bbox.bbox_id for bbox in self.bboxes]
+        if bbox_id not in bboxes_ids:
+            self.bboxes.append(Bbox(bbox_id, top, left, width, height))
+        else:
+            raise ValueError("Frame with id: {} already has a Bbox with id: {}".format(self.frame_id, bbox_id))
+
+    def add_label_to_bbox(self, bbox_id: int, category: str, confidence: float):
+        bboxes = {bbox.id: bbox for bbox in self.bboxes}
+        if bbox_id in bboxes.keys():
+            res = bboxes.get(bbox_id)
+            res.add_label(category, confidence)
+        else:
+            raise ValueError('the bbox with id: {} does not exists!'.format(bbox_id))
+
+
+class BboxToJsonLogger(BaseJsonLogger):
+    """
+    ُ This module is designed to automate the task of logging jsons. An example json is used
+    to show the contents of json file shortly
+    Example:
+          {
+          "video_details": {
+            "frame_width": 1920,
+            "frame_height": 1080,
+            "frame_rate": 20,
+            "video_name": "/home/gpu/codes/MSD/pedestrian_2/project/public/camera1.avi"
+          },
+          "frames": [
+            {
+              "frame_id": 329,
+              "timestamp": 3365.1254
+              "bboxes": [
+                {
+                  "labels": [
+                    {
+                      "category": "pedestrian",
+                      "confidence": 0.9
+                    }
+                  ],
+                  "bbox_id": 0,
+                  "top": 1257,
+                  "left": 138,
+                  "width": 68,
+                  "height": 109
+                }
+              ]
+            }],
+
+    Attributes:
+        frames (dict): It's a dictionary that maps each frame_id to json attributes.
+        video_details (dict): information about video file.
+        top_k_labels (int): shows the allowed number of labels
+        start_time (datetime object): we use it to automate the json output by time.
+
+    Args:
+        top_k_labels (int): shows the allowed number of labels
+
+    """
+
+    def __init__(self, top_k_labels: int = 1):
+        self.frames = {}
+        self.video_details = self.video_details = dict(frame_width=None, frame_height=None, frame_rate=None,
+                                                       video_name=None)
+        self.top_k_labels = top_k_labels
+        self.start_time = datetime.now()
+
+    def set_top_k(self, value):
+        self.top_k_labels = value
+
+    def frame_exists(self, frame_id: int) -> bool:
+        """
+        Args:
+            frame_id (int):
+
+        Returns:
+            bool: true if frame_id is recognized
+        """
+        return frame_id in self.frames.keys()
+
+    def add_frame(self, frame_id: int, timestamp: float = None) -> None:
+        """
+        Args:
+            frame_id (int):
+            timestamp (float): opencv captured frame time property
+
+        Raises:
+             ValueError: if frame_id would not exist in class frames attribute
+
+        Returns:
+            None
+
+        """
+        if not self.frame_exists(frame_id):
+            self.frames[frame_id] = Frame(frame_id, timestamp)
+        else:
+            raise ValueError("Frame id: {} already exists".format(frame_id))
+
+    def bbox_exists(self, frame_id: int, bbox_id: int) -> bool:
+        """
+        Args:
+            frame_id:
+            bbox_id:
+
+        Returns:
+            bool: if bbox exists in frame bboxes list
+        """
+        bboxes = []
+        if self.frame_exists(frame_id=frame_id):
+            bboxes = [bbox.bbox_id for bbox in self.frames[frame_id].bboxes]
+        return bbox_id in bboxes
+
+    def find_bbox(self, frame_id: int, bbox_id: int):
+        """
+
+        Args:
+            frame_id:
+            bbox_id:
+
+        Returns:
+            bbox_id (int):
+
+        Raises:
+            ValueError: if bbox_id does not exist in the bbox list of specific frame.
+        """
+        if not self.bbox_exists(frame_id, bbox_id):
+            raise ValueError("frame with id: {} does not contain bbox with id: {}".format(frame_id, bbox_id))
+        bboxes = {bbox.bbox_id: bbox for bbox in self.frames[frame_id].bboxes}
+        return bboxes.get(bbox_id)
+
+    def add_bbox_to_frame(self, frame_id: int, bbox_id: int, top: int, left: int, width: int, height: int) -> None:
+        """
+
+        Args:
+            frame_id (int):
+            bbox_id (int):
+            top (int):
+            left (int):
+            width (int):
+            height (int):
+
+        Returns:
+            None
+
+        Raises:
+            ValueError: if bbox_id already exist in frame information with frame_id
+            ValueError: if frame_id does not exist in frames attribute
+        """
+        if self.frame_exists(frame_id):
+            frame = self.frames[frame_id]
+            if not self.bbox_exists(frame_id, bbox_id):
+                frame.add_bbox(bbox_id, top, left, width, height)
+            else:
+                raise ValueError(
+                    "frame with frame_id: {} already contains the bbox with id: {} ".format(frame_id, bbox_id))
+        else:
+            raise ValueError("frame with frame_id: {} does not exist".format(frame_id))
+
+    def add_label_to_bbox(self, frame_id: int, bbox_id: int, category: str, confidence: float):
+        """
+        Args:
+            frame_id:
+            bbox_id:
+            category:
+            confidence: the confidence value returned from yolo detection
+
+        Returns:
+            None
+
+        Raises:
+            ValueError: if labels quota (top_k_labels) exceeds.
+        """
+        bbox = self.find_bbox(frame_id, bbox_id)
+        if not bbox.labels_full(self.top_k_labels):
+            bbox.add_label(category, confidence)
+        else:
+            raise ValueError("labels in frame_id: {}, bbox_id: {} is fulled".format(frame_id, bbox_id))
+
+    def add_video_details(self, frame_width: int = None, frame_height: int = None, frame_rate: int = None,
+                          video_name: str = None):
+        self.video_details['frame_width'] = frame_width
+        self.video_details['frame_height'] = frame_height
+        self.video_details['frame_rate'] = frame_rate
+        self.video_details['video_name'] = video_name
+
+    def output(self):
+        output = {'video_details': self.video_details}
+        result = list(self.frames.values())
+        output['frames'] = [item.dic() for item in result]
+        return output
+
+    def json_output(self, output_name):
+        """
+        Args:
+            output_name:
+
+        Returns:
+            None
+
+        Notes:
+            It creates the json output with `output_name` name.
+        """
+        if not output_name.endswith('.json'):
+            output_name += '.json'
+        with open(output_name, 'w') as file:
+            json.dump(self.output(), file)
+        file.close()
+
+    def set_start(self):
+        self.start_time = datetime.now()
+
+    def schedule_output_by_time(self, output_dir=JsonMeta.PATH_TO_SAVE, hours: int = 0, minutes: int = 0,
+                                seconds: int = 60) -> None:
+        """
+        Notes:
+            Creates folder and then periodically stores the jsons on that address.
+
+        Args:
+            output_dir (str): the directory where output files will be stored
+            hours (int):
+            minutes (int):
+            seconds (int):
+
+        Returns:
+            None
+
+        """
+        end = datetime.now()
+        interval = 0
+        interval += abs(min([hours, JsonMeta.HOURS]) * 3600)
+        interval += abs(min([minutes, JsonMeta.MINUTES]) * 60)
+        interval += abs(min([seconds, JsonMeta.SECONDS]))
+        diff = (end - self.start_time).seconds
+
+        if diff > interval:
+            output_name = self.start_time.strftime('%Y-%m-%d %H-%M-%S') + '.json'
+            if not exists(output_dir):
+                makedirs(output_dir)
+            output = join(output_dir, output_name)
+            self.json_output(output_name=output)
+            self.frames = {}
+            self.start_time = datetime.now()
+
+    def schedule_output_by_frames(self, frames_quota, frame_counter, output_dir=JsonMeta.PATH_TO_SAVE):
+        """
+        saves as the number of frames quota increases higher.
+        :param frames_quota:
+        :param frame_counter:
+        :param output_dir:
+        :return:
+        """
+        pass
+
+    def flush(self, output_dir):
+        """
+        Notes:
+            We use this function to output jsons whenever possible.
+            like the time that we exit the while loop of opencv.
+
+        Args:
+            output_dir:
+
+        Returns:
+            None
+
+        """
+        filename = self.start_time.strftime('%Y-%m-%d %H-%M-%S') + '-remaining.json'
+        output = join(output_dir, filename)
+        self.json_output(output_name=output)

deep_sort/utils/log.py

@@ -0,0 +1,17 @@
+import logging
+
+
+def get_logger(name='root'):
+    formatter = logging.Formatter(
+        # fmt='%(asctime)s [%(levelname)s]: %(filename)s(%(funcName)s:%(lineno)s) >> %(message)s')
+        fmt='%(asctime)s [%(levelname)s]: %(message)s', datefmt='%Y-%m-%d %H:%M:%S')
+
+    handler = logging.StreamHandler()
+    handler.setFormatter(formatter)
+
+    logger = logging.getLogger(name)
+    logger.setLevel(logging.INFO)
+    logger.addHandler(handler)
+    return logger
+
+

deep_sort/utils/parser.py

@@ -0,0 +1,38 @@
+import os
+import yaml
+from easydict import EasyDict as edict
+
+class YamlParser(edict):
+    """
+    This is yaml parser based on EasyDict.
+    """
+    def __init__(self, cfg_dict=None, config_file=None):
+        if cfg_dict is None:
+            cfg_dict = {}
+
+        if config_file is not None:
+            assert(os.path.isfile(config_file))
+            with open(config_file, 'r') as fo:
+                cfg_dict.update(yaml.load(fo.read()))
+
+        super(YamlParser, self).__init__(cfg_dict)
+
+    
+    def merge_from_file(self, config_file):
+        with open(config_file, 'r') as fo:
+            self.update(yaml.load(fo.read()))
+
+    
+    def merge_from_dict(self, config_dict):
+        self.update(config_dict)
+
+
+def get_config(config_file=None):
+    return YamlParser(config_file=config_file)
+
+
+if __name__ == "__main__":
+    cfg = YamlParser(config_file="../configs/yolov3.yaml")
+    cfg.merge_from_file("../configs/deep_sort.yaml")
+
+    import ipdb; ipdb.set_trace()

deep_sort/utils/tools.py

@@ -0,0 +1,39 @@
+from functools import wraps
+from time import time
+
+
+def is_video(ext: str):
+    """
+    Returns true if ext exists in
+    allowed_exts for video files.
+
+    Args:
+        ext:
+
+    Returns:
+
+    """
+
+    allowed_exts = ('.mp4', '.webm', '.ogg', '.avi', '.wmv', '.mkv', '.3gp')
+    return any((ext.endswith(x) for x in allowed_exts))
+
+
+def tik_tok(func):
+    """
+    keep track of time for each process.
+    Args:
+        func:
+
+    Returns:
+
+    """
+    @wraps(func)
+    def _time_it(*args, **kwargs):
+        start = time()
+        try:
+            return func(*args, **kwargs)
+        finally:
+            end_ = time()
+            print("time: {:.03f}s, fps: {:.03f}".format(end_ - start, 1 / (end_ - start)))
+
+    return _time_it

demo.py

@@ -0,0 +1,54 @@
+from AIDetector_pytorch import Detector
+import imutils
+import cv2
+
+def main():
+
+    func_status = {}
+    func_status['headpose'] = None
+    
+    name = 'demo'
+
+    det = Detector()
+    cap = cv2.VideoCapture('E:/视频/行人监控/test01.mp4')
+    fps = int(cap.get(5))
+    print('fps:', fps)
+    t = int(1000/fps)
+
+    size = None
+    videoWriter = None
+
+    while True:
+
+        # try:
+        _, im = cap.read()
+        if im is None:
+            break
+        
+        result = det.feedCap(im, func_status)
+        result = result['frame']
+        result = imutils.resize(result, height=500)
+        if videoWriter is None:
+            fourcc = cv2.VideoWriter_fourcc(
+                'm', 'p', '4', 'v')  # opencv3.0
+            videoWriter = cv2.VideoWriter(
+                'result.mp4', fourcc, fps, (result.shape[1], result.shape[0]))
+
+        videoWriter.write(result)
+        cv2.imshow(name, result)
+        cv2.waitKey(t)
+
+        if cv2.getWindowProperty(name, cv2.WND_PROP_AUTOSIZE) < 1:
+            # 点x退出
+            break
+        # except Exception as e:
+        #     print(e)
+        #     break
+
+    cap.release()
+    videoWriter.release()
+    cv2.destroyAllWindows()
+
+if __name__ == '__main__':
+    
+    main()

models/common.py

@@ -0,0 +1,114 @@
+# This file contains modules common to various models
+
+import math
+
+import torch
+import torch.nn as nn
+
+from utils.general import non_max_suppression
+
+def autopad(k, p=None):  # kernel, padding
+    # Pad to 'same'
+    if p is None:
+        p = k // 2 if isinstance(k, int) else [x // 2 for x in k]  # auto-pad
+    return p
+
+
+def DWConv(c1, c2, k=1, s=1, act=True):
+    # Depthwise convolution
+    return Conv(c1, c2, k, s, g=math.gcd(c1, c2), act=act)
+
+
+class Conv(nn.Module):
+    # Standard convolution
+    def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True):  # ch_in, ch_out, kernel, stride, padding, groups
+        super(Conv, self).__init__()
+        self.conv = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=False)
+        self.bn = nn.BatchNorm2d(c2)
+        self.act = nn.Hardswish() if act else nn.Identity()
+
+    def forward(self, x):
+        return self.act(self.bn(self.conv(x)))
+
+    def fuseforward(self, x):
+        return self.act(self.conv(x))
+
+
+class Bottleneck(nn.Module):
+    # Standard bottleneck
+    def __init__(self, c1, c2, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, shortcut, groups, expansion
+        super(Bottleneck, self).__init__()
+        c_ = int(c2 * e)  # hidden channels
+        self.cv1 = Conv(c1, c_, 1, 1)
+        self.cv2 = Conv(c_, c2, 3, 1, g=g)
+        self.add = shortcut and c1 == c2
+
+    def forward(self, x):
+        return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x))
+
+
+class BottleneckCSP(nn.Module):
+    # CSP Bottleneck https://github.com/WongKinYiu/CrossStagePartialNetworks
+    def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, number, shortcut, groups, expansion
+        super(BottleneckCSP, self).__init__()
+        c_ = int(c2 * e)  # hidden channels
+        self.cv1 = Conv(c1, c_, 1, 1)
+        self.cv2 = nn.Conv2d(c1, c_, 1, 1, bias=False)
+        self.cv3 = nn.Conv2d(c_, c_, 1, 1, bias=False)
+        self.cv4 = Conv(2 * c_, c2, 1, 1)
+        self.bn = nn.BatchNorm2d(2 * c_)  # applied to cat(cv2, cv3)
+        self.act = nn.LeakyReLU(0.1, inplace=True)
+        self.m = nn.Sequential(*[Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)])
+
+    def forward(self, x):
+        y1 = self.cv3(self.m(self.cv1(x)))
+        y2 = self.cv2(x)
+        return self.cv4(self.act(self.bn(torch.cat((y1, y2), dim=1))))
+
+
+class SPP(nn.Module):
+    # Spatial pyramid pooling layer used in YOLOv3-SPP
+    def __init__(self, c1, c2, k=(5, 9, 13)):
+        super(SPP, self).__init__()
+        c_ = c1 // 2  # hidden channels
+        self.cv1 = Conv(c1, c_, 1, 1)
+        self.cv2 = Conv(c_ * (len(k) + 1), c2, 1, 1)
+        self.m = nn.ModuleList([nn.MaxPool2d(kernel_size=x, stride=1, padding=x // 2) for x in k])
+
+    def forward(self, x):
+        x = self.cv1(x)
+        return self.cv2(torch.cat([x] + [m(x) for m in self.m], 1))
+
+
+class Focus(nn.Module):
+    # Focus wh information into c-space
+    def __init__(self, c1, c2, k=1, s=1, p=None, g=1, act=True):  # ch_in, ch_out, kernel, stride, padding, groups
+        super(Focus, self).__init__()
+        self.conv = Conv(c1 * 4, c2, k, s, p, g, act)
+
+    def forward(self, x):  # x(b,c,w,h) -> y(b,4c,w/2,h/2)
+        return self.conv(torch.cat([x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]], 1))
+
+
+class Concat(nn.Module):
+    # Concatenate a list of tensors along dimension
+    def __init__(self, dimension=1):
+        super(Concat, self).__init__()
+        self.d = dimension
+
+    def forward(self, x):
+        return torch.cat(x, self.d)
+
+
+class NMS(nn.Module):
+    # Non-Maximum Suppression (NMS) module
+    conf = 0.25  # confidence threshold
+    iou = 0.45  # IoU threshold
+    classes = None  # (optional list) filter by class
+
+    def __init__(self):
+        super(NMS, self).__init__()
+
+    def forward(self, x):
+        return non_max_suppression(x[0], conf_thres=self.conf, iou_thres=self.iou, classes=self.classes)
+

models/experimental.py

@@ -0,0 +1,152 @@
+# This file contains experimental modules
+
+import numpy as np
+import torch
+import torch.nn as nn
+
+from models.common import Conv, DWConv
+from utils.google_utils import attempt_download
+
+
+class CrossConv(nn.Module):
+    # Cross Convolution Downsample
+    def __init__(self, c1, c2, k=3, s=1, g=1, e=1.0, shortcut=False):
+        # ch_in, ch_out, kernel, stride, groups, expansion, shortcut
+        super(CrossConv, self).__init__()
+        c_ = int(c2 * e)  # hidden channels
+        self.cv1 = Conv(c1, c_, (1, k), (1, s))
+        self.cv2 = Conv(c_, c2, (k, 1), (s, 1), g=g)
+        self.add = shortcut and c1 == c2
+
+    def forward(self, x):
+        return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x))
+
+
+class C3(nn.Module):
+    # Cross Convolution CSP
+    def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, number, shortcut, groups, expansion
+        super(C3, self).__init__()
+        c_ = int(c2 * e)  # hidden channels
+        self.cv1 = Conv(c1, c_, 1, 1)
+        self.cv2 = nn.Conv2d(c1, c_, 1, 1, bias=False)
+        self.cv3 = nn.Conv2d(c_, c_, 1, 1, bias=False)
+        self.cv4 = Conv(2 * c_, c2, 1, 1)
+        self.bn = nn.BatchNorm2d(2 * c_)  # applied to cat(cv2, cv3)
+        self.act = nn.LeakyReLU(0.1, inplace=True)
+        self.m = nn.Sequential(*[CrossConv(c_, c_, 3, 1, g, 1.0, shortcut) for _ in range(n)])
+
+    def forward(self, x):
+        y1 = self.cv3(self.m(self.cv1(x)))
+        y2 = self.cv2(x)
+        return self.cv4(self.act(self.bn(torch.cat((y1, y2), dim=1))))
+
+
+class Sum(nn.Module):
+    # Weighted sum of 2 or more layers https://arxiv.org/abs/1911.09070
+    def __init__(self, n, weight=False):  # n: number of inputs
+        super(Sum, self).__init__()
+        self.weight = weight  # apply weights boolean
+        self.iter = range(n - 1)  # iter object
+        if weight:
+            self.w = nn.Parameter(-torch.arange(1., n) / 2, requires_grad=True)  # layer weights
+
+    def forward(self, x):
+        y = x[0]  # no weight
+        if self.weight:
+            w = torch.sigmoid(self.w) * 2
+            for i in self.iter:
+                y = y + x[i + 1] * w[i]
+        else:
+            for i in self.iter:
+                y = y + x[i + 1]
+        return y
+
+
+class GhostConv(nn.Module):
+    # Ghost Convolution https://github.com/huawei-noah/ghostnet
+    def __init__(self, c1, c2, k=1, s=1, g=1, act=True):  # ch_in, ch_out, kernel, stride, groups
+        super(GhostConv, self).__init__()
+        c_ = c2 // 2  # hidden channels
+        self.cv1 = Conv(c1, c_, k, s, None, g, act)
+        self.cv2 = Conv(c_, c_, 5, 1, None, c_, act)
+
+    def forward(self, x):
+        y = self.cv1(x)
+        return torch.cat([y, self.cv2(y)], 1)
+
+
+class GhostBottleneck(nn.Module):
+    # Ghost Bottleneck https://github.com/huawei-noah/ghostnet
+    def __init__(self, c1, c2, k, s):
+        super(GhostBottleneck, self).__init__()
+        c_ = c2 // 2
+        self.conv = nn.Sequential(GhostConv(c1, c_, 1, 1),  # pw
+                                  DWConv(c_, c_, k, s, act=False) if s == 2 else nn.Identity(),  # dw
+                                  GhostConv(c_, c2, 1, 1, act=False))  # pw-linear
+        self.shortcut = nn.Sequential(DWConv(c1, c1, k, s, act=False),
+                                      Conv(c1, c2, 1, 1, act=False)) if s == 2 else nn.Identity()
+
+    def forward(self, x):
+        return self.conv(x) + self.shortcut(x)
+
+
+class MixConv2d(nn.Module):
+    # Mixed Depthwise Conv https://arxiv.org/abs/1907.09595
+    def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True):
+        super(MixConv2d, self).__init__()
+        groups = len(k)
+        if equal_ch:  # equal c_ per group
+            i = torch.linspace(0, groups - 1E-6, c2).floor()  # c2 indices
+            c_ = [(i == g).sum() for g in range(groups)]  # intermediate channels
+        else:  # equal weight.numel() per group
+            b = [c2] + [0] * groups
+            a = np.eye(groups + 1, groups, k=-1)
+            a -= np.roll(a, 1, axis=1)
+            a *= np.array(k) ** 2
+            a[0] = 1
+            c_ = np.linalg.lstsq(a, b, rcond=None)[0].round()  # solve for equal weight indices, ax = b
+
+        self.m = nn.ModuleList([nn.Conv2d(c1, int(c_[g]), k[g], s, k[g] // 2, bias=False) for g in range(groups)])
+        self.bn = nn.BatchNorm2d(c2)
+        self.act = nn.LeakyReLU(0.1, inplace=True)
+
+    def forward(self, x):
+        return x + self.act(self.bn(torch.cat([m(x) for m in self.m], 1)))
+
+
+class Ensemble(nn.ModuleList):
+    # Ensemble of models
+    def __init__(self):
+        super(Ensemble, self).__init__()
+
+    def forward(self, x, augment=False):
+        y = []
+        for module in self:
+            y.append(module(x, augment)[0])
+        # y = torch.stack(y).max(0)[0]  # max ensemble
+        # y = torch.cat(y, 1)  # nms ensemble
+        y = torch.stack(y).mean(0)  # mean ensemble
+        return y, None  # inference, train output
+
+
+def attempt_load(weights, map_location=None):
+    # Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a
+    model = Ensemble()
+    for w in weights if isinstance(weights, list) else [weights]:
+        attempt_download(w)
+        model.append(torch.load(w, map_location=map_location)['model'].float().fuse().eval())  # load FP32 model
+
+    # Compatibility updates
+    for m in model.modules():
+        if type(m) in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6]:
+            m.inplace = True  # pytorch 1.7.0 compatibility
+        elif type(m) is Conv:
+            m._non_persistent_buffers_set = set()  # pytorch 1.6.0 compatibility
+
+    if len(model) == 1:
+        return model[-1]  # return model
+    else:
+        print('Ensemble created with %s\n' % weights)
+        for k in ['names', 'stride']:
+            setattr(model, k, getattr(model[-1], k))
+        return model  # return ensemble

models/yolo.py

@@ -0,0 +1,244 @@
+import math
+import sys
+from copy import deepcopy
+from pathlib import Path
+
+sys.path.append('./')  # to run '$ python *.py' files in subdirectories
+
+import torch
+import torch.nn as nn
+
+from models.common import Conv, Bottleneck, SPP, DWConv, Focus, BottleneckCSP, Concat, NMS
+from models.experimental import MixConv2d, CrossConv, C3
+from utils.autoanchor import check_anchor_order
+from utils.general import make_divisible
+from utils.torch_utils import time_synchronized, fuse_conv_and_bn, model_info, scale_img, initialize_weights
+
+thop = None
+
+
+class Detect(nn.Module):
+    stride = None  # strides computed during build
+    export = False  # onnx export
+
+    def __init__(self, nc=80, anchors=(), ch=()):  # detection layer
+        super(Detect, self).__init__()
+        self.nc = nc  # number of classes
+        self.no = nc + 5  # number of outputs per anchor
+        self.nl = len(anchors)  # number of detection layers
+        self.na = len(anchors[0]) // 2  # number of anchors
+        self.grid = [torch.zeros(1)] * self.nl  # init grid
+        a = torch.tensor(anchors).float().view(self.nl, -1, 2)
+        self.register_buffer('anchors', a)  # shape(nl,na,2)
+        self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2))  # shape(nl,1,na,1,1,2)
+        self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch)  # output conv
+
+    def forward(self, x):
+        # x = x.copy()  # for profiling
+        z = []  # inference output
+        self.training |= self.export
+        for i in range(self.nl):
+            x[i] = self.m[i](x[i])  # conv
+            bs, _, ny, nx = x[i].shape  # x(bs,255,20,20) to x(bs,3,20,20,85)
+            x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
+
+            if not self.training:  # inference
+                if self.grid[i].shape[2:4] != x[i].shape[2:4]:
+                    self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
+
+                y = x[i].sigmoid()
+                y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i].to(x[i].device)) * self.stride[i]  # xy
+                y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i]  # wh
+                z.append(y.view(bs, -1, self.no))
+
+        return x if self.training else (torch.cat(z, 1), x)
+
+    @staticmethod
+    def _make_grid(nx=20, ny=20):
+        yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
+        return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
+
+
+class Model(nn.Module):
+    def __init__(self, cfg='yolov5m.yaml', ch=3, nc=None):  # model, input channels, number of classes
+        super(Model, self).__init__()
+        if isinstance(cfg, dict):
+            self.yaml = cfg  # model dict
+        else:  # is *.yaml
+            import yaml  # for torch hub
+            self.yaml_file = Path(cfg).name
+            with open(cfg) as f:
+                self.yaml = yaml.load(f, Loader=yaml.FullLoader)  # model dict
+
+        # Define model
+        if nc and nc != self.yaml['nc']:
+            self.yaml['nc'] = nc  # override yaml value
+        self.model, self.save = parse_model(deepcopy(self.yaml), ch=[ch])  # model, savelist, ch_out
+        # print([x.shape for x in self.forward(torch.zeros(1, ch, 64, 64))])
+
+        # Build strides, anchors
+        m = self.model[-1]  # Detect()
+        if isinstance(m, Detect):
+            s = 128  # 2x min stride
+            m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))])  # forward
+            m.anchors /= m.stride.view(-1, 1, 1)
+            check_anchor_order(m)
+            self.stride = m.stride
+            self._initialize_biases()  # only run once
+            # print('Strides: %s' % m.stride.tolist())
+
+        # Init weights, biases
+        initialize_weights(self)
+        self.info()
+
+    def forward(self, x, augment=False, profile=False):
+        if augment:
+            img_size = x.shape[-2:]  # height, width
+            s = [1, 0.83, 0.67]  # scales
+            f = [None, 3, None]  # flips (2-ud, 3-lr)
+            y = []  # outputs
+            for si, fi in zip(s, f):
+                xi = scale_img(x.flip(fi) if fi else x, si)
+                yi = self.forward_once(xi)[0]  # forward
+                # cv2.imwrite('img%g.jpg' % s, 255 * xi[0].numpy().transpose((1, 2, 0))[:, :, ::-1])  # save
+                yi[..., :4] /= si  # de-scale
+                if fi == 2:
+                    yi[..., 1] = img_size[0] - yi[..., 1]  # de-flip ud
+                elif fi == 3:
+                    yi[..., 0] = img_size[1] - yi[..., 0]  # de-flip lr
+                y.append(yi)
+            return torch.cat(y, 1), None  # augmented inference, train
+        else:
+            return self.forward_once(x, profile)  # single-scale inference, train
+
+    def forward_once(self, x, profile=False):
+        y, dt = [], []  # outputs
+        for m in self.model:
+            if m.f != -1:  # if not from previous layer
+                x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f]  # from earlier layers
+
+            if profile:
+                o = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2 if thop else 0  # FLOPS
+                t = time_synchronized()
+                for _ in range(10):
+                    _ = m(x)
+                dt.append((time_synchronized() - t) * 100)
+                print('%10.1f%10.0f%10.1fms %-40s' % (o, m.np, dt[-1], m.type))
+
+            x = m(x)  # run
+            y.append(x if m.i in self.save else None)  # save output
+
+        if profile:
+            print('%.1fms total' % sum(dt))
+        return x
+
+    def _initialize_biases(self, cf=None):  # initialize biases into Detect(), cf is class frequency
+        # https://arxiv.org/abs/1708.02002 section 3.3
+        # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
+        m = self.model[-1]  # Detect() module
+        for mi, s in zip(m.m, m.stride):  # from
+            b = mi.bias.view(m.na, -1)  # conv.bias(255) to (3,85)
+            b[:, 4] += math.log(8 / (640 / s) ** 2)  # obj (8 objects per 640 image)
+            b[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum())  # cls
+            mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
+
+    def _print_biases(self):
+        m = self.model[-1]  # Detect() module
+        for mi in m.m:  # from
+            b = mi.bias.detach().view(m.na, -1).T  # conv.bias(255) to (3,85)
+            print(('%6g Conv2d.bias:' + '%10.3g' * 6) % (mi.weight.shape[1], *b[:5].mean(1).tolist(), b[5:].mean()))
+
+    # def _print_weights(self):
+    #     for m in self.model.modules():
+    #         if type(m) is Bottleneck:
+    #             print('%10.3g' % (m.w.detach().sigmoid() * 2))  # shortcut weights
+
+    def fuse(self):  # fuse model Conv2d() + BatchNorm2d() layers
+        print('Fusing layers... ')
+        for m in self.model.modules():
+            if type(m) is Conv and hasattr(m, 'bn'):
+                m.conv = fuse_conv_and_bn(m.conv, m.bn)  # update conv
+                delattr(m, 'bn')  # remove batchnorm
+                m.forward = m.fuseforward  # update forward
+        self.info()
+        return self
+
+    def nms(self, mode=True):  # add or remove NMS module
+        present = type(self.model[-1]) is NMS  # last layer is NMS
+        if mode and not present:
+            print('Adding NMS... ')
+            m = NMS()  # module
+            m.f = -1  # from
+            m.i = self.model[-1].i + 1  # index
+            self.model.add_module(name='%s' % m.i, module=m)  # add
+            self.eval()
+        elif not mode and present:
+            print('Removing NMS... ')
+            self.model = self.model[:-1]  # remove
+        return self
+
+    def info(self, verbose=False, img_size=640):  # print model information
+        model_info(self, verbose, img_size)
+
+
+def parse_model(d, ch):  # model_dict, input_channels(3)
+    anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple']
+    na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors  # number of anchors
+    no = na * (nc + 5)  # number of outputs = anchors * (classes + 5)
+
+    layers, save, c2 = [], [], ch[-1]  # layers, savelist, ch out
+    for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']):  # from, number, module, args
+        m = eval(m) if isinstance(m, str) else m  # eval strings
+        for j, a in enumerate(args):
+            try:
+                args[j] = eval(a) if isinstance(a, str) else a  # eval strings
+            except:
+                pass
+
+        n = max(round(n * gd), 1) if n > 1 else n  # depth gain
+        if m in [Conv, Bottleneck, SPP, DWConv, MixConv2d, Focus, CrossConv, BottleneckCSP, C3]:
+            c1, c2 = ch[f], args[0]
+
+            # Normal
+            # if i > 0 and args[0] != no:  # channel expansion factor
+            #     ex = 1.75  # exponential (default 2.0)
+            #     e = math.log(c2 / ch[1]) / math.log(2)
+            #     c2 = int(ch[1] * ex ** e)
+            # if m != Focus:
+
+            c2 = make_divisible(c2 * gw, 8) if c2 != no else c2
+
+            # Experimental
+            # if i > 0 and args[0] != no:  # channel expansion factor
+            #     ex = 1 + gw  # exponential (default 2.0)
+            #     ch1 = 32  # ch[1]
+            #     e = math.log(c2 / ch1) / math.log(2)  # level 1-n
+            #     c2 = int(ch1 * ex ** e)
+            # if m != Focus:
+            #     c2 = make_divisible(c2, 8) if c2 != no else c2
+
+            args = [c1, c2, *args[1:]]
+            if m in [BottleneckCSP, C3]:
+                args.insert(2, n)
+                n = 1
+        elif m is nn.BatchNorm2d:
+            args = [ch[f]]
+        elif m is Concat:
+            c2 = sum([ch[-1 if x == -1 else x + 1] for x in f])
+        elif m is Detect:
+            args.append([ch[x + 1] for x in f])
+            if isinstance(args[1], int):  # number of anchors
+                args[1] = [list(range(args[1] * 2))] * len(f)
+        else:
+            c2 = ch[f]
+
+        m_ = nn.Sequential(*[m(*args) for _ in range(n)]) if n > 1 else m(*args)  # module
+        t = str(m)[8:-2].replace('__main__.', '')  # module type
+        np = sum([x.numel() for x in m_.parameters()])  # number params
+        m_.i, m_.f, m_.type, m_.np = i, f, t, np  # attach index, 'from' index, type, number params
+        save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1)  # append to savelist
+        layers.append(m_)
+        ch.append(c2)
+    return nn.Sequential(*layers), sorted(save)
+
+

models/yolov5m.yaml

@@ -0,0 +1,48 @@
+# parameters
+nc: 8  # number of classes
+depth_multiple: 0.67  # model depth multiple
+width_multiple: 0.75  # layer channel multiple
+
+# anchors
+anchors:
+  - [10,13, 16,30, 33,23]  # P3/8
+  - [30,61, 62,45, 59,119]  # P4/16
+  - [116,90, 156,198, 373,326]  # P5/32
+
+# YOLOv5 backbone
+backbone:
+  # [from, number, module, args]
+  [[-1, 1, Focus, [64, 3]],  # 0-P1/2
+   [-1, 1, Conv, [128, 3, 2]],  # 1-P2/4
+   [-1, 3, BottleneckCSP, [128]],
+   [-1, 1, Conv, [256, 3, 2]],  # 3-P3/8
+   [-1, 9, BottleneckCSP, [256]],
+   [-1, 1, Conv, [512, 3, 2]],  # 5-P4/16
+   [-1, 9, BottleneckCSP, [512]],
+   [-1, 1, Conv, [1024, 3, 2]],  # 7-P5/32
+   [-1, 1, SPP, [1024, [5, 9, 13]]],
+   [-1, 3, BottleneckCSP, [1024, False]],  # 9
+  ]
+
+# YOLOv5 head
+head:
+  [[-1, 1, Conv, [512, 1, 1]],
+   [-1, 1, nn.Upsample, [None, 2, 'nearest']],
+   [[-1, 6], 1, Concat, [1]],  # cat backbone P4
+   [-1, 3, BottleneckCSP, [512, False]],  # 13
+
+   [-1, 1, Conv, [256, 1, 1]],
+   [-1, 1, nn.Upsample, [None, 2, 'nearest']],
+   [[-1, 4], 1, Concat, [1]],  # cat backbone P3
+   [-1, 3, BottleneckCSP, [256, False]],  # 17 (P3/8-small)
+
+   [-1, 1, Conv, [256, 3, 2]],
+   [[-1, 14], 1, Concat, [1]],  # cat head P4
+   [-1, 3, BottleneckCSP, [512, False]],  # 20 (P4/16-medium)
+
+   [-1, 1, Conv, [512, 3, 2]],
+   [[-1, 10], 1, Concat, [1]],  # cat head P5
+   [-1, 3, BottleneckCSP, [1024, False]],  # 23 (P5/32-large)
+
+   [[17, 20, 23], 1, Detect, [nc, anchors]],  # Detect(P3, P4, P5)
+  ]

tracker.py

@@ -0,0 +1,74 @@
+from deep_sort.utils.parser import get_config
+from deep_sort.deep_sort import DeepSort
+import torch
+import cv2
+
+palette = (2 ** 11 - 1, 2 ** 15 - 1, 2 ** 20 - 1)
+cfg = get_config()
+cfg.merge_from_file("deep_sort/configs/deep_sort.yaml")
+deepsort = DeepSort(cfg.DEEPSORT.REID_CKPT,
+                    max_dist=cfg.DEEPSORT.MAX_DIST, min_confidence=cfg.DEEPSORT.MIN_CONFIDENCE,
+                    nms_max_overlap=cfg.DEEPSORT.NMS_MAX_OVERLAP, max_iou_distance=cfg.DEEPSORT.MAX_IOU_DISTANCE,
+                    max_age=cfg.DEEPSORT.MAX_AGE, n_init=cfg.DEEPSORT.N_INIT, nn_budget=cfg.DEEPSORT.NN_BUDGET,
+                    use_cuda=True)
+
+
+def plot_bboxes(image, bboxes, line_thickness=None):
+    # Plots one bounding box on image img
+    tl = line_thickness or round(
+        0.002 * (image.shape[0] + image.shape[1]) / 2) + 1  # line/font thickness
+    for (x1, y1, x2, y2, cls_id, pos_id) in bboxes:
+        if cls_id in ['smoke', 'phone', 'eat']:
+            color = (0, 0, 255)
+        else:
+            color = (0, 255, 0)
+        if cls_id == 'eat':
+            cls_id = 'eat-drink'
+        c1, c2 = (x1, y1), (x2, y2)
+        cv2.rectangle(image, c1, c2, color, thickness=tl, lineType=cv2.LINE_AA)
+        tf = max(tl - 1, 1)  # font thickness
+        t_size = cv2.getTextSize(cls_id, 0, fontScale=tl / 3, thickness=tf)[0]
+        c2 = c1[0] + t_size[0], c1[1] - t_size[1] - 3
+        cv2.rectangle(image, c1, c2, color, -1, cv2.LINE_AA)  # filled
+        cv2.putText(image, '{} ID-{}'.format(cls_id, pos_id), (c1[0], c1[1] - 2), 0, tl / 3,
+                    [225, 255, 255], thickness=tf, lineType=cv2.LINE_AA)
+
+    return image
+
+
+def update_tracker(target_detector, image):
+
+        new_faces = []
+        _, bboxes = target_detector.detect(image)
+
+        bbox_xywh = []
+        confs = []
+
+        # Adapt detections to deep sort input format
+        for x1, y1, x2, y2, _, conf in bboxes:
+            
+            obj = [
+                int((x1+x2)/2), int((y1+y2)/2),
+                x2-x1, y2-y1
+            ]
+            bbox_xywh.append(obj)
+            confs.append(conf)
+
+        xywhs = torch.Tensor(bbox_xywh)
+        confss = torch.Tensor(confs)
+
+        # Pass detections to deepsort
+        outputs = deepsort.update(xywhs, confss, image)
+
+
+        bboxes2draw = []
+        face_bboxes = []
+        for value in list(outputs):
+            x1,y1,x2,y2,track_id = value
+            bboxes2draw.append(
+                (x1, y1, x2, y2, '', track_id)
+            )
+
+        image = plot_bboxes(image, bboxes2draw)
+
+        return image, new_faces, face_bboxes

utils/BaseDetector.py

@@ -0,0 +1,50 @@
+from tracker import update_tracker
+import cv2
+
+
+class baseDet(object):
+
+    def __init__(self):
+
+        self.img_size = 640
+        self.threshold = 0.3
+        self.stride = 1
+
+    def build_config(self):
+
+        self.faceTracker = {}
+        self.faceClasses = {}
+        self.faceLocation1 = {}
+        self.faceLocation2 = {}
+        self.frameCounter = 0
+        self.currentCarID = 0
+        self.recorded = []
+
+        self.font = cv2.FONT_HERSHEY_SIMPLEX
+
+    def feedCap(self, im, func_status):
+
+        retDict = {
+            'frame': None,
+            'faces': None,
+            'list_of_ids': None,
+            'face_bboxes': []
+        }
+        self.frameCounter += 1
+
+        im, faces, face_bboxes = update_tracker(self, im)
+
+        retDict['frame'] = im
+        retDict['faces'] = faces
+        retDict['face_bboxes'] = face_bboxes
+
+        return retDict
+
+    def init_model(self):
+        raise EOFError("Undefined model type.")
+
+    def preprocess(self):
+        raise EOFError("Undefined model type.")
+
+    def detect(self):
+        raise EOFError("Undefined model type.")

utils/activations.py

@@ -0,0 +1,72 @@
+# Activation functions
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+# Swish https://arxiv.org/pdf/1905.02244.pdf ---------------------------------------------------------------------------
+class Swish(nn.Module):  #
+    @staticmethod
+    def forward(x):
+        return x * torch.sigmoid(x)
+
+
+class Hardswish(nn.Module):  # export-friendly version of nn.Hardswish()
+    @staticmethod
+    def forward(x):
+        # return x * F.hardsigmoid(x)  # for torchscript and CoreML
+        return x * F.hardtanh(x + 3, 0., 6.) / 6.  # for torchscript, CoreML and ONNX
+
+
+class MemoryEfficientSwish(nn.Module):
+    class F(torch.autograd.Function):
+        @staticmethod
+        def forward(ctx, x):
+            ctx.save_for_backward(x)
+            return x * torch.sigmoid(x)
+
+        @staticmethod
+        def backward(ctx, grad_output):
+            x = ctx.saved_tensors[0]
+            sx = torch.sigmoid(x)
+            return grad_output * (sx * (1 + x * (1 - sx)))
+
+    def forward(self, x):
+        return self.F.apply(x)
+
+
+# Mish https://github.com/digantamisra98/Mish --------------------------------------------------------------------------
+class Mish(nn.Module):
+    @staticmethod
+    def forward(x):
+        return x * F.softplus(x).tanh()
+
+
+class MemoryEfficientMish(nn.Module):
+    class F(torch.autograd.Function):
+        @staticmethod
+        def forward(ctx, x):
+            ctx.save_for_backward(x)
+            return x.mul(torch.tanh(F.softplus(x)))  # x * tanh(ln(1 + exp(x)))
+
+        @staticmethod
+        def backward(ctx, grad_output):
+            x = ctx.saved_tensors[0]
+            sx = torch.sigmoid(x)
+            fx = F.softplus(x).tanh()
+            return grad_output * (fx + x * sx * (1 - fx * fx))
+
+    def forward(self, x):
+        return self.F.apply(x)
+
+
+# FReLU https://arxiv.org/abs/2007.11824 -------------------------------------------------------------------------------
+class FReLU(nn.Module):
+    def __init__(self, c1, k=3):  # ch_in, kernel
+        super().__init__()
+        self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1)
+        self.bn = nn.BatchNorm2d(c1)
+
+    def forward(self, x):
+        return torch.max(x, self.bn(self.conv(x)))

utils/autoanchor.py

@@ -0,0 +1,151 @@
+# Auto-anchor utils
+
+import numpy as np
+import torch
+import yaml
+from scipy.cluster.vq import kmeans
+from tqdm import tqdm
+
+
+def check_anchor_order(m):
+    # Check anchor order against stride order for YOLOv5 Detect() module m, and correct if necessary
+    a = m.anchor_grid.prod(-1).view(-1)  # anchor area
+    da = a[-1] - a[0]  # delta a
+    ds = m.stride[-1] - m.stride[0]  # delta s
+    if da.sign() != ds.sign():  # same order
+        print('Reversing anchor order')
+        m.anchors[:] = m.anchors.flip(0)
+        m.anchor_grid[:] = m.anchor_grid.flip(0)
+
+
+def check_anchors(dataset, model, thr=4.0, imgsz=640):
+    # Check anchor fit to data, recompute if necessary
+    print('\nAnalyzing anchors... ', end='')
+    m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1]  # Detect()
+    shapes = imgsz * dataset.shapes / dataset.shapes.max(1, keepdims=True)
+    scale = np.random.uniform(0.9, 1.1, size=(shapes.shape[0], 1))  # augment scale
+    wh = torch.tensor(np.concatenate([l[:, 3:5] * s for s, l in zip(shapes * scale, dataset.labels)])).float()  # wh
+
+    def metric(k):  # compute metric
+        r = wh[:, None] / k[None]
+        x = torch.min(r, 1. / r).min(2)[0]  # ratio metric
+        best = x.max(1)[0]  # best_x
+        aat = (x > 1. / thr).float().sum(1).mean()  # anchors above threshold
+        bpr = (best > 1. / thr).float().mean()  # best possible recall
+        return bpr, aat
+
+    bpr, aat = metric(m.anchor_grid.clone().cpu().view(-1, 2))
+    print('anchors/target = %.2f, Best Possible Recall (BPR) = %.4f' % (aat, bpr), end='')
+    if bpr < 0.98:  # threshold to recompute
+        print('. Attempting to improve anchors, please wait...')
+        na = m.anchor_grid.numel() // 2  # number of anchors
+        new_anchors = kmean_anchors(dataset, n=na, img_size=imgsz, thr=thr, gen=1000, verbose=False)
+        new_bpr = metric(new_anchors.reshape(-1, 2))[0]
+        if new_bpr > bpr:  # replace anchors
+            new_anchors = torch.tensor(new_anchors, device=m.anchors.device).type_as(m.anchors)
+            m.anchor_grid[:] = new_anchors.clone().view_as(m.anchor_grid)  # for inference
+            m.anchors[:] = new_anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1)  # loss
+            check_anchor_order(m)
+            print('New anchors saved to model. Update model *.yaml to use these anchors in the future.')
+        else:
+            print('Original anchors better than new anchors. Proceeding with original anchors.')
+    print('')  # newline
+
+
+def kmean_anchors(path='./data/coco128.yaml', n=9, img_size=640, thr=4.0, gen=1000, verbose=True):
+    """ Creates kmeans-evolved anchors from training dataset
+
+        Arguments:
+            path: path to dataset *.yaml, or a loaded dataset
+            n: number of anchors
+            img_size: image size used for training
+            thr: anchor-label wh ratio threshold hyperparameter hyp['anchor_t'] used for training, default=4.0
+            gen: generations to evolve anchors using genetic algorithm
+            verbose: print all results
+
+        Return:
+            k: kmeans evolved anchors
+
+        Usage:
+            from utils.autoanchor import *; _ = kmean_anchors()
+    """
+    thr = 1. / thr
+
+    def metric(k, wh):  # compute metrics
+        r = wh[:, None] / k[None]
+        x = torch.min(r, 1. / r).min(2)[0]  # ratio metric
+        # x = wh_iou(wh, torch.tensor(k))  # iou metric
+        return x, x.max(1)[0]  # x, best_x
+
+    def anchor_fitness(k):  # mutation fitness
+        _, best = metric(torch.tensor(k, dtype=torch.float32), wh)
+        return (best * (best > thr).float()).mean()  # fitness
+
+    def print_results(k):
+        k = k[np.argsort(k.prod(1))]  # sort small to large
+        x, best = metric(k, wh0)
+        bpr, aat = (best > thr).float().mean(), (x > thr).float().mean() * n  # best possible recall, anch > thr
+        print('thr=%.2f: %.4f best possible recall, %.2f anchors past thr' % (thr, bpr, aat))
+        print('n=%g, img_size=%s, metric_all=%.3f/%.3f-mean/best, past_thr=%.3f-mean: ' %
+              (n, img_size, x.mean(), best.mean(), x[x > thr].mean()), end='')
+        for i, x in enumerate(k):
+            print('%i,%i' % (round(x[0]), round(x[1])), end=',  ' if i < len(k) - 1 else '\n')  # use in *.cfg
+        return k
+
+    if isinstance(path, str):  # *.yaml file
+        with open(path) as f:
+            data_dict = yaml.load(f, Loader=yaml.FullLoader)  # model dict
+        from utils.datasets import LoadImagesAndLabels
+        dataset = LoadImagesAndLabels(data_dict['train'], augment=True, rect=True)
+    else:
+        dataset = path  # dataset
+
+    # Get label wh
+    shapes = img_size * dataset.shapes / dataset.shapes.max(1, keepdims=True)
+    wh0 = np.concatenate([l[:, 3:5] * s for s, l in zip(shapes, dataset.labels)])  # wh
+
+    # Filter
+    i = (wh0 < 3.0).any(1).sum()
+    if i:
+        print('WARNING: Extremely small objects found. '
+              '%g of %g labels are < 3 pixels in width or height.' % (i, len(wh0)))
+    wh = wh0[(wh0 >= 2.0).any(1)]  # filter > 2 pixels
+
+    # Kmeans calculation
+    print('Running kmeans for %g anchors on %g points...' % (n, len(wh)))
+    s = wh.std(0)  # sigmas for whitening
+    k, dist = kmeans(wh / s, n, iter=30)  # points, mean distance
+    k *= s
+    wh = torch.tensor(wh, dtype=torch.float32)  # filtered
+    wh0 = torch.tensor(wh0, dtype=torch.float32)  # unfiltered
+    k = print_results(k)
+
+    # Plot
+    # k, d = [None] * 20, [None] * 20
+    # for i in tqdm(range(1, 21)):
+    #     k[i-1], d[i-1] = kmeans(wh / s, i)  # points, mean distance
+    # fig, ax = plt.subplots(1, 2, figsize=(14, 7), tight_layout=True)
+    # ax = ax.ravel()
+    # ax[0].plot(np.arange(1, 21), np.array(d) ** 2, marker='.')
+    # fig, ax = plt.subplots(1, 2, figsize=(14, 7))  # plot wh
+    # ax[0].hist(wh[wh[:, 0]<100, 0],400)
+    # ax[1].hist(wh[wh[:, 1]<100, 1],400)
+    # fig.savefig('wh.png', dpi=200)
+
+    # Evolve
+    npr = np.random
+    f, sh, mp, s = anchor_fitness(k), k.shape, 0.9, 0.1  # fitness, generations, mutation prob, sigma
+    pbar = tqdm(range(gen), desc='Evolving anchors with Genetic Algorithm')  # progress bar
+    for _ in pbar:
+        v = np.ones(sh)
+        while (v == 1).all():  # mutate until a change occurs (prevent duplicates)
+            v = ((npr.random(sh) < mp) * npr.random() * npr.randn(*sh) * s + 1).clip(0.3, 3.0)
+        kg = (k.copy() * v).clip(min=2.0)
+        fg = anchor_fitness(kg)
+        if fg > f:
+            f, k = fg, kg.copy()
+            pbar.desc = 'Evolving anchors with Genetic Algorithm: fitness = %.4f' % f
+            if verbose:
+                print_results(k)
+
+    return print_results(k)

utils/general.py

@@ -0,0 +1,502 @@
+# General utils
+
+import glob
+import os
+import platform
+import random
+import re
+import subprocess
+import time
+from pathlib import Path
+
+import cv2
+import math
+import numpy as np
+import torch
+import torchvision
+import yaml
+
+from utils.google_utils import gsutil_getsize
+from utils.metrics import fitness
+from utils.torch_utils import init_torch_seeds
+
+# Settings
+torch.set_printoptions(linewidth=320, precision=5, profile='long')
+# format short g, %precision=5
+np.set_printoptions(linewidth=320, formatter={'float_kind': '{:11.5g}'.format})
+# prevent OpenCV from multithreading (incompatible with PyTorch DataLoader)
+cv2.setNumThreads(0)
+
+
+def init_seeds(seed=0):
+    random.seed(seed)
+    np.random.seed(seed)
+    init_torch_seeds(seed)
+
+
+def get_latest_run(search_dir='.'):
+    # Return path to most recent 'last.pt' in /runs (i.e. to --resume from)
+    last_list = glob.glob(f'{search_dir}/**/last*.pt', recursive=True)
+    return max(last_list, key=os.path.getctime) if last_list else ''
+
+
+def check_git_status():
+    # Suggest 'git pull' if repo is out of date
+    if platform.system() in ['Linux', 'Darwin'] and not os.path.isfile('/.dockerenv'):
+        s = subprocess.check_output(
+            'if [ -d .git ]; then git fetch && git status -uno; fi', shell=True).decode('utf-8')
+        if 'Your branch is behind' in s:
+            print(s[s.find('Your branch is behind'):s.find('\n\n')] + '\n')
+
+
+def check_img_size(img_size, s=32):
+    # Verify img_size is a multiple of stride s
+    new_size = make_divisible(img_size, int(s))  # ceil gs-multiple
+    if new_size != img_size:
+        print('WARNING: --img-size %g must be multiple of max stride %g, updating to %g' %
+              (img_size, s, new_size))
+    return new_size
+
+
+def check_file(file):
+    # Search for file if not found
+    if os.path.isfile(file) or file == '':
+        return file
+    else:
+        files = glob.glob('./**/' + file, recursive=True)  # find file
+        assert len(files), 'File Not Found: %s' % file  # assert file was found
+        assert len(files) == 1, "Multiple files match '%s', specify exact path: %s" % (
+            file, files)  # assert unique
+        return files[0]  # return file
+
+
+def check_dataset(dict):
+    # Download dataset if not found locally
+    val, s = dict.get('val'), dict.get('download')
+    if val and len(val):
+        val = [Path(x).resolve()
+               for x in (val if isinstance(val, list) else [val])]  # val path
+        if not all(x.exists() for x in val):
+            print('\nWARNING: Dataset not found, nonexistent paths: %s' %
+                  [str(x) for x in val if not x.exists()])
+            if s and len(s):  # download script
+                print('Downloading %s ...' % s)
+                if s.startswith('http') and s.endswith('.zip'):  # URL
+                    f = Path(s).name  # filename
+                    torch.hub.download_url_to_file(s, f)
+                    r = os.system('unzip -q %s -d ../ && rm %s' %
+                                  (f, f))  # unzip
+                else:  # bash script
+                    r = os.system(s)
+                print('Dataset autodownload %s\n' % ('success' if r ==
+                                                     0 else 'failure'))  # analyze return value
+            else:
+                raise Exception('Dataset not found.')
+
+
+def make_divisible(x, divisor):
+    # Returns x evenly divisible by divisor
+    return math.ceil(x / divisor) * divisor
+
+
+def labels_to_class_weights(labels, nc=80):
+    # Get class weights (inverse frequency) from training labels
+    if labels[0] is None:  # no labels loaded
+        return torch.Tensor()
+
+    labels = np.concatenate(labels, 0)  # labels.shape = (866643, 5) for COCO
+    classes = labels[:, 0].astype(np.int)  # labels = [class xywh]
+    weights = np.bincount(classes, minlength=nc)  # occurrences per class
+
+    # Prepend gridpoint count (for uCE training)
+    # gpi = ((320 / 32 * np.array([1, 2, 4])) ** 2 * 3).sum()  # gridpoints per image
+    # weights = np.hstack([gpi * len(labels)  - weights.sum() * 9, weights * 9]) ** 0.5  # prepend gridpoints to start
+
+    weights[weights == 0] = 1  # replace empty bins with 1
+    weights = 1 / weights  # number of targets per class
+    weights /= weights.sum()  # normalize
+    return torch.from_numpy(weights)
+
+
+def labels_to_image_weights(labels, nc=80, class_weights=np.ones(80)):
+    # Produces image weights based on class_weights and image contents
+    class_counts = np.array(
+        [np.bincount(x[:, 0].astype(np.int), minlength=nc) for x in labels])
+    image_weights = (class_weights.reshape(1, nc) * class_counts).sum(1)
+    # index = random.choices(range(n), weights=image_weights, k=1)  # weight image sample
+    return image_weights
+
+
+def coco80_to_coco91_class():  # converts 80-index (val2014) to 91-index (paper)
+    # https://tech.amikelive.com/node-718/what-object-categories-labels-are-in-coco-dataset/
+    # a = np.loadtxt('data/coco.names', dtype='str', delimiter='\n')
+    # b = np.loadtxt('data/coco_paper.names', dtype='str', delimiter='\n')
+    # x1 = [list(a[i] == b).index(True) + 1 for i in range(80)]  # darknet to coco
+    # x2 = [list(b[i] == a).index(True) if any(b[i] == a) else None for i in range(91)]  # coco to darknet
+    x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34,
+         35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63,
+         64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88, 89, 90]
+    return x
+
+
+def xyxy2xywh(x):
+    # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] where xy1=top-left, xy2=bottom-right
+    y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
+    y[:, 0] = (x[:, 0] + x[:, 2]) / 2  # x center
+    y[:, 1] = (x[:, 1] + x[:, 3]) / 2  # y center
+    y[:, 2] = x[:, 2] - x[:, 0]  # width
+    y[:, 3] = x[:, 3] - x[:, 1]  # height
+    return y
+
+
+def xywh2xyxy(x):
+    # Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
+    y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
+    y[:, 0] = x[:, 0] - x[:, 2] / 2  # top left x
+    y[:, 1] = x[:, 1] - x[:, 3] / 2  # top left y
+    y[:, 2] = x[:, 0] + x[:, 2] / 2  # bottom right x
+    y[:, 3] = x[:, 1] + x[:, 3] / 2  # bottom right y
+    return y
+
+
+def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None):
+    # Rescale coords (xyxy) from img1_shape to img0_shape
+    if ratio_pad is None:  # calculate from img0_shape
+        gain = min(img1_shape[0] / img0_shape[0],
+                   img1_shape[1] / img0_shape[1])  # gain  = old / new
+        pad = (img1_shape[1] - img0_shape[1] * gain) / \
+            2, (img1_shape[0] - img0_shape[0] * gain) / 2  # wh padding
+    else:
+        gain = ratio_pad[0][0]
+        pad = ratio_pad[1]
+
+    coords[:, [0, 2]] -= pad[0]  # x padding
+    coords[:, [1, 3]] -= pad[1]  # y padding
+    coords[:, :4] /= gain
+    clip_coords(coords, img0_shape)
+    return coords
+
+
+def clip_coords(boxes, img_shape):
+    # Clip bounding xyxy bounding boxes to image shape (height, width)
+    boxes[:, 0].clamp_(0, img_shape[1])  # x1
+    boxes[:, 1].clamp_(0, img_shape[0])  # y1
+    boxes[:, 2].clamp_(0, img_shape[1])  # x2
+    boxes[:, 3].clamp_(0, img_shape[0])  # y2
+
+
+def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-9):
+    # Returns the IoU of box1 to box2. box1 is 4, box2 is nx4
+    box2 = box2.T
+
+    # Get the coordinates of bounding boxes
+    if x1y1x2y2:  # x1, y1, x2, y2 = box1
+        b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3]
+        b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3]
+    else:  # transform from xywh to xyxy
+        b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2
+        b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2
+        b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2
+        b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2
+
+    # Intersection area
+    inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \
+            (torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0)
+
+    # Union Area
+    w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
+    w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
+    union = w1 * h1 + w2 * h2 - inter + eps
+
+    iou = inter / union
+    if GIoU or DIoU or CIoU:
+        # convex (smallest enclosing box) width
+        cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1)
+        ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1)  # convex height
+        if CIoU or DIoU:  # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1
+            c2 = cw ** 2 + ch ** 2 + eps  # convex diagonal squared
+            rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 +
+                    (b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4  # center distance squared
+            if DIoU:
+                return iou - rho2 / c2  # DIoU
+            elif CIoU:  # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47
+                v = (4 / math.pi ** 2) * \
+                    torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2)
+                with torch.no_grad():
+                    alpha = v / ((1 + eps) - iou + v)
+                return iou - (rho2 / c2 + v * alpha)  # CIoU
+        else:  # GIoU https://arxiv.org/pdf/1902.09630.pdf
+            c_area = cw * ch + eps  # convex area
+            return iou - (c_area - union) / c_area  # GIoU
+    else:
+        return iou  # IoU
+
+
+def box_iou(box1, box2):
+    # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
+    """
+    Return intersection-over-union (Jaccard index) of boxes.
+    Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
+    Arguments:
+        box1 (Tensor[N, 4])
+        box2 (Tensor[M, 4])
+    Returns:
+        iou (Tensor[N, M]): the NxM matrix containing the pairwise
+            IoU values for every element in boxes1 and boxes2
+    """
+
+    def box_area(box):
+        # box = 4xn
+        return (box[2] - box[0]) * (box[3] - box[1])
+
+    area1 = box_area(box1.T)
+    area2 = box_area(box2.T)
+
+    # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2)
+    inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) -
+             torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
+    # iou = inter / (area1 + area2 - inter)
+    return inter / (area1[:, None] + area2 - inter)
+
+
+def wh_iou(wh1, wh2):
+    # Returns the nxm IoU matrix. wh1 is nx2, wh2 is mx2
+    wh1 = wh1[:, None]  # [N,1,2]
+    wh2 = wh2[None]  # [1,M,2]
+    inter = torch.min(wh1, wh2).prod(2)  # [N,M]
+    # iou = inter / (area1 + area2 - inter)
+    return inter / (wh1.prod(2) + wh2.prod(2) - inter)
+
+
+def non_max_suppression(prediction, conf_thres=0.1, iou_thres=0.6, classes=None, agnostic=False, labels=()):
+    """Performs Non-Maximum Suppression (NMS) on inference results
+
+    Returns:
+         detections with shape: nx6 (x1, y1, x2, y2, conf, cls)
+    """
+
+    nc = prediction[0].shape[1] - 5  # number of classes
+    xc = prediction[..., 4] > conf_thres  # candidates
+
+    # Settings
+    # (pixels) minimum and maximum box width and height
+    min_wh, max_wh = 2, 4096
+    max_det = 300  # maximum number of detections per image
+    time_limit = 10.0  # seconds to quit after
+    redundant = True  # require redundant detections
+    multi_label = nc > 1  # multiple labels per box (adds 0.5ms/img)
+    merge = False  # use merge-NMS
+
+    t = time.time()
+    output = [torch.zeros(0, 6)] * prediction.shape[0]
+    for xi, x in enumerate(prediction):  # image index, image inference
+        # Apply constraints
+        # x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0  # width-height
+        x = x[xc[xi]]  # confidence
+
+        # Cat apriori labels if autolabelling
+        if labels and len(labels[xi]):
+            l = labels[xi]
+            v = torch.zeros((len(l), nc + 5), device=x.device)
+            v[:, :4] = l[:, 1:5]  # box
+            v[:, 4] = 1.0  # conf
+            v[range(len(l)), l[:, 0].long() + 5] = 1.0  # cls
+            x = torch.cat((x, v), 0)
+
+        # If none remain process next image
+        if not x.shape[0]:
+            continue
+
+        # Compute conf
+        x[:, 5:] *= x[:, 4:5]  # conf = obj_conf * cls_conf
+
+        # Box (center x, center y, width, height) to (x1, y1, x2, y2)
+        box = xywh2xyxy(x[:, :4])
+
+        # Detections matrix nx6 (xyxy, conf, cls)
+        if multi_label:
+            i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T
+            x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1)
+        else:  # best class only
+            conf, j = x[:, 5:].max(1, keepdim=True)
+            x = torch.cat((box, conf, j.float()), 1)[
+                conf.view(-1) > conf_thres]
+
+        # Filter by class
+        if classes:
+            x = x[(x[:, 5:6] == torch.tensor(classes, device=x.device)).any(1)]
+
+        # Apply finite constraint
+        # if not torch.isfinite(x).all():
+        #     x = x[torch.isfinite(x).all(1)]
+
+        # If none remain process next image
+        n = x.shape[0]  # number of boxes
+        if not n:
+            continue
+
+        # Sort by confidence
+        # x = x[x[:, 4].argsort(descending=True)]
+
+        # Batched NMS
+        c = x[:, 5:6] * (0 if agnostic else max_wh)  # classes
+        # boxes (offset by class), scores
+        boxes, scores = x[:, :4] + c, x[:, 4]
+        i = torchvision.ops.nms(boxes, scores, iou_thres)  # NMS
+        if i.shape[0] > max_det:  # limit detections
+            i = i[:max_det]
+        if merge and (1 < n < 3E3):  # Merge NMS (boxes merged using weighted mean)
+            # update boxes as boxes(i,4) = weights(i,n) * boxes(n,4)
+            iou = box_iou(boxes[i], boxes) > iou_thres  # iou matrix
+            weights = iou * scores[None]  # box weights
+            x[i, :4] = torch.mm(weights, x[:, :4]).float(
+            ) / weights.sum(1, keepdim=True)  # merged boxes
+            if redundant:
+                i = i[iou.sum(1) > 1]  # require redundancy
+
+        output[xi] = x[i]
+        if (time.time() - t) > time_limit:
+            break  # time limit exceeded
+
+    return output
+
+
+# from utils.general import *; strip_optimizer()
+def strip_optimizer(f='weights/best.pt', s=''):
+    # Strip optimizer from 'f' to finalize training, optionally save as 's'
+    x = torch.load(f, map_location=torch.device('cpu'))
+    x['optimizer'] = None
+    x['training_results'] = None
+    x['epoch'] = -1
+    x['model'].half()  # to FP16
+    for p in x['model'].parameters():
+        p.requires_grad = False
+    torch.save(x, s or f)
+    mb = os.path.getsize(s or f) / 1E6  # filesize
+    print('Optimizer stripped from %s,%s %.1fMB' %
+          (f, (' saved as %s,' % s) if s else '', mb))
+
+
+def print_mutation(hyp, results, yaml_file='hyp_evolved.yaml', bucket=''):
+    # Print mutation results to evolve.txt (for use with train.py --evolve)
+    a = '%10s' * len(hyp) % tuple(hyp.keys())  # hyperparam keys
+    b = '%10.3g' * len(hyp) % tuple(hyp.values())  # hyperparam values
+    # results (P, R, mAP@0.5, mAP@0.5:0.95, val_losses x 3)
+    c = '%10.4g' * len(results) % results
+    print('\n%s\n%s\nEvolved fitness: %s\n' % (a, b, c))
+
+    if bucket:
+        url = 'gs://%s/evolve.txt' % bucket
+        if gsutil_getsize(url) > (os.path.getsize('evolve.txt') if os.path.exists('evolve.txt') else 0):
+            # download evolve.txt if larger than local
+            os.system('gsutil cp %s .' % url)
+
+    with open('evolve.txt', 'a') as f:  # append result
+        f.write(c + b + '\n')
+    x = np.unique(np.loadtxt('evolve.txt', ndmin=2),
+                  axis=0)  # load unique rows
+    x = x[np.argsort(-fitness(x))]  # sort
+    np.savetxt('evolve.txt', x, '%10.3g')  # save sort by fitness
+
+    # Save yaml
+    for i, k in enumerate(hyp.keys()):
+        hyp[k] = float(x[0, i + 7])
+    with open(yaml_file, 'w') as f:
+        results = tuple(x[0, :7])
+        # results (P, R, mAP@0.5, mAP@0.5:0.95, val_losses x 3)
+        c = '%10.4g' * len(results) % results
+        f.write('# Hyperparameter Evolution Results\n# Generations: %g\n# Metrics: ' % len(
+            x) + c + '\n\n')
+        yaml.dump(hyp, f, sort_keys=False)
+
+    if bucket:
+        os.system('gsutil cp evolve.txt %s gs://%s' %
+                  (yaml_file, bucket))  # upload
+
+
+def apply_classifier(x, model, img, im0):
+    # applies a second stage classifier to yolo outputs
+    im0 = [im0] if isinstance(im0, np.ndarray) else im0
+    for i, d in enumerate(x):  # per image
+        if d is not None and len(d):
+            d = d.clone()
+
+            # Reshape and pad cutouts
+            b = xyxy2xywh(d[:, :4])  # boxes
+            b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1)  # rectangle to square
+            b[:, 2:] = b[:, 2:] * 1.3 + 30  # pad
+            d[:, :4] = xywh2xyxy(b).long()
+
+            # Rescale boxes from img_size to im0 size
+            scale_coords(img.shape[2:], d[:, :4], im0[i].shape)
+
+            # Classes
+            pred_cls1 = d[:, 5].long()
+            ims = []
+            for j, a in enumerate(d):  # per item
+                cutout = im0[i][int(a[1]):int(a[3]), int(a[0]):int(a[2])]
+                im = cv2.resize(cutout, (224, 224))  # BGR
+                # cv2.imwrite('test%i.jpg' % j, cutout)
+
+                # BGR to RGB, to 3x416x416
+                im = im[:, :, ::-1].transpose(2, 0, 1)
+                im = np.ascontiguousarray(
+                    im, dtype=np.float32)  # uint8 to float32
+                im /= 255.0  # 0 - 255 to 0.0 - 1.0
+                ims.append(im)
+
+            pred_cls2 = model(torch.Tensor(ims).to(d.device)
+                              ).argmax(1)  # classifier prediction
+            # retain matching class detections
+            x[i] = x[i][pred_cls1 == pred_cls2]
+
+    return x
+
+
+def increment_path(path, exist_ok=True, sep=''):
+    # Increment path, i.e. runs/exp --> runs/exp{sep}0, runs/exp{sep}1 etc.
+    path = Path(path)  # os-agnostic
+    if (path.exists() and exist_ok) or (not path.exists()):
+        return str(path)
+    else:
+        dirs = glob.glob(f"{path}{sep}*")  # similar paths
+        matches = [re.search(rf"%s{sep}(\d+)" % path.stem, d) for d in dirs]
+        i = [int(m.groups()[0]) for m in matches if m]  # indices
+        n = max(i) + 1 if i else 2  # increment number
+        return f"{path}{sep}{n}"  # update path
+
+
+def letterbox(img, new_shape=(640, 640), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True):
+    shape = img.shape[:2]  # current shape [height, width]
+    if isinstance(new_shape, int):
+        new_shape = (new_shape, new_shape)
+
+    # Scale ratio (new / old)
+    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
+    if not scaleup:  # only scale down, do not scale up (for better test mAP)
+        r = min(r, 1.0)
+
+    # Compute padding
+    ratio = r, r  # width, height ratios
+    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
+    dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - \
+        new_unpad[1]  # wh padding
+    if auto:  # minimum rectangle
+        dw, dh = np.mod(dw, 32), np.mod(dh, 32)  # wh padding
+    elif scaleFill:  # stretch
+        dw, dh = 0.0, 0.0
+        new_unpad = (new_shape[1], new_shape[0])
+        ratio = new_shape[1] / shape[1], new_shape[0] / \
+            shape[0]  # width, height ratios
+
+    dw /= 2  # divide padding into 2 sides
+    dh /= 2
+
+    if shape[::-1] != new_unpad:  # resize
+        img = cv2.resize(img, new_unpad, interpolation=cv2.INTER_LINEAR)
+    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
+    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
+    img = cv2.copyMakeBorder(
+        img, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
+    return img, ratio, (dw, dh)
+

utils/google_utils.py

@@ -0,0 +1,122 @@
+# Google utils: https://cloud.google.com/storage/docs/reference/libraries
+
+import os
+import platform
+import subprocess
+import time
+from pathlib import Path
+
+import torch
+
+
+def gsutil_getsize(url=''):
+    # gs://bucket/file size https://cloud.google.com/storage/docs/gsutil/commands/du
+    s = subprocess.check_output('gsutil du %s' % url, shell=True).decode('utf-8')
+    return eval(s.split(' ')[0]) if len(s) else 0  # bytes
+
+
+def attempt_download(weights):
+    # Attempt to download pretrained weights if not found locally
+    weights = weights.strip().replace("'", '')
+    file = Path(weights).name.lower()
+
+    msg = weights + ' missing, try downloading from https://github.com/ultralytics/yolov5/releases/'
+    models = ['yolov5s.pt', 'yolov5m.pt', 'yolov5l.pt', 'yolov5x.pt']  # available models
+    redundant = False  # offer second download option
+
+    if file in models and not os.path.isfile(weights):
+        # Google Drive
+        # d = {'yolov5s.pt': '1R5T6rIyy3lLwgFXNms8whc-387H0tMQO',
+        #      'yolov5m.pt': '1vobuEExpWQVpXExsJ2w-Mbf3HJjWkQJr',
+        #      'yolov5l.pt': '1hrlqD1Wdei7UT4OgT785BEk1JwnSvNEV',
+        #      'yolov5x.pt': '1mM8aZJlWTxOg7BZJvNUMrTnA2AbeCVzS'}
+        # r = gdrive_download(id=d[file], name=weights) if file in d else 1
+        # if r == 0 and os.path.exists(weights) and os.path.getsize(weights) > 1E6:  # check
+        #    return
+
+        try:  # GitHub
+            url = 'https://github.com/ultralytics/yolov5/releases/download/v3.1/' + file
+            print('Downloading %s to %s...' % (url, weights))
+            torch.hub.download_url_to_file(url, weights)
+            assert os.path.exists(weights) and os.path.getsize(weights) > 1E6  # check
+        except Exception as e:  # GCP
+            print('Download error: %s' % e)
+            assert redundant, 'No secondary mirror'
+            url = 'https://storage.googleapis.com/ultralytics/yolov5/ckpt/' + file
+            print('Downloading %s to %s...' % (url, weights))
+            r = os.system('curl -L %s -o %s' % (url, weights))  # torch.hub.download_url_to_file(url, weights)
+        finally:
+            if not (os.path.exists(weights) and os.path.getsize(weights) > 1E6):  # check
+                os.remove(weights) if os.path.exists(weights) else None  # remove partial downloads
+                print('ERROR: Download failure: %s' % msg)
+            print('')
+            return
+
+
+def gdrive_download(id='1n_oKgR81BJtqk75b00eAjdv03qVCQn2f', name='coco128.zip'):
+    # Downloads a file from Google Drive. from utils.google_utils import *; gdrive_download()
+    t = time.time()
+
+    print('Downloading https://drive.google.com/uc?export=download&id=%s as %s... ' % (id, name), end='')
+    os.remove(name) if os.path.exists(name) else None  # remove existing
+    os.remove('cookie') if os.path.exists('cookie') else None
+
+    # Attempt file download
+    out = "NUL" if platform.system() == "Windows" else "/dev/null"
+    os.system('curl -c ./cookie -s -L "drive.google.com/uc?export=download&id=%s" > %s ' % (id, out))
+    if os.path.exists('cookie'):  # large file
+        s = 'curl -Lb ./cookie "drive.google.com/uc?export=download&confirm=%s&id=%s" -o %s' % (get_token(), id, name)
+    else:  # small file
+        s = 'curl -s -L -o %s "drive.google.com/uc?export=download&id=%s"' % (name, id)
+    r = os.system(s)  # execute, capture return
+    os.remove('cookie') if os.path.exists('cookie') else None
+
+    # Error check
+    if r != 0:
+        os.remove(name) if os.path.exists(name) else None  # remove partial
+        print('Download error ')  # raise Exception('Download error')
+        return r
+
+    # Unzip if archive
+    if name.endswith('.zip'):
+        print('unzipping... ', end='')
+        os.system('unzip -q %s' % name)  # unzip
+        os.remove(name)  # remove zip to free space
+
+    print('Done (%.1fs)' % (time.time() - t))
+    return r
+
+
+def get_token(cookie="./cookie"):
+    with open(cookie) as f:
+        for line in f:
+            if "download" in line:
+                return line.split()[-1]
+    return ""
+
+# def upload_blob(bucket_name, source_file_name, destination_blob_name):
+#     # Uploads a file to a bucket
+#     # https://cloud.google.com/storage/docs/uploading-objects#storage-upload-object-python
+#
+#     storage_client = storage.Client()
+#     bucket = storage_client.get_bucket(bucket_name)
+#     blob = bucket.blob(destination_blob_name)
+#
+#     blob.upload_from_filename(source_file_name)
+#
+#     print('File {} uploaded to {}.'.format(
+#         source_file_name,
+#         destination_blob_name))
+#
+#
+# def download_blob(bucket_name, source_blob_name, destination_file_name):
+#     # Uploads a blob from a bucket
+#     storage_client = storage.Client()
+#     bucket = storage_client.get_bucket(bucket_name)
+#     blob = bucket.blob(source_blob_name)
+#
+#     blob.download_to_filename(destination_file_name)
+#
+#     print('Blob {} downloaded to {}.'.format(
+#         source_blob_name,
+#         destination_file_name))

utils/metrics.py

@@ -0,0 +1,201 @@
+# Model validation metrics
+
+from pathlib import Path
+
+import matplotlib.pyplot as plt
+import numpy as np
+import torch
+
+from . import general
+
+
+def fitness(x):
+    # Model fitness as a weighted combination of metrics
+    w = [0.0, 0.0, 0.1, 0.9]  # weights for [P, R, mAP@0.5, mAP@0.5:0.95]
+    return (x[:, :4] * w).sum(1)
+
+
+def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='precision-recall_curve.png', names=[]):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.
+    # Arguments
+        tp:  True positives (nparray, nx1 or nx10).
+        conf:  Objectness value from 0-1 (nparray).
+        pred_cls:  Predicted object classes (nparray).
+        target_cls:  True object classes (nparray).
+        plot:  Plot precision-recall curve at mAP@0.5
+        save_dir:  Plot save directory
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Sort by objectness
+    i = np.argsort(-conf)
+    tp, conf, pred_cls = tp[i], conf[i], pred_cls[i]
+
+    # Find unique classes
+    unique_classes = np.unique(target_cls)
+
+    # Create Precision-Recall curve and compute AP for each class
+    px, py = np.linspace(0, 1, 1000), []  # for plotting
+    pr_score = 0.1  # score to evaluate P and R https://github.com/ultralytics/yolov3/issues/898
+    s = [unique_classes.shape[0], tp.shape[1]]  # number class, number iou thresholds (i.e. 10 for mAP0.5...0.95)
+    ap, p, r = np.zeros(s), np.zeros(s), np.zeros(s)
+    for ci, c in enumerate(unique_classes):
+        i = pred_cls == c
+        n_l = (target_cls == c).sum()  # number of labels
+        n_p = i.sum()  # number of predictions
+
+        if n_p == 0 or n_l == 0:
+            continue
+        else:
+            # Accumulate FPs and TPs
+            fpc = (1 - tp[i]).cumsum(0)
+            tpc = tp[i].cumsum(0)
+
+            # Recall
+            recall = tpc / (n_l + 1e-16)  # recall curve
+            r[ci] = np.interp(-pr_score, -conf[i], recall[:, 0])  # r at pr_score, negative x, xp because xp decreases
+
+            # Precision
+            precision = tpc / (tpc + fpc)  # precision curve
+            p[ci] = np.interp(-pr_score, -conf[i], precision[:, 0])  # p at pr_score
+
+            # AP from recall-precision curve
+            for j in range(tp.shape[1]):
+                ap[ci, j], mpre, mrec = compute_ap(recall[:, j], precision[:, j])
+                if plot and (j == 0):
+                    py.append(np.interp(px, mrec, mpre))  # precision at mAP@0.5
+
+    # Compute F1 score (harmonic mean of precision and recall)
+    f1 = 2 * p * r / (p + r + 1e-16)
+
+    if plot:
+        plot_pr_curve(px, py, ap, save_dir, names)
+
+    return p, r, ap, f1, unique_classes.astype('int32')
+
+
+def compute_ap(recall, precision):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rbgirshick/py-faster-rcnn.
+    # Arguments
+        recall:    The recall curve (list).
+        precision: The precision curve (list).
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Append sentinel values to beginning and end
+    mrec = recall  # np.concatenate(([0.], recall, [recall[-1] + 1E-3]))
+    mpre = precision  # np.concatenate(([0.], precision, [0.]))
+
+    # Compute the precision envelope
+    mpre = np.flip(np.maximum.accumulate(np.flip(mpre)))
+
+    # Integrate area under curve
+    method = 'interp'  # methods: 'continuous', 'interp'
+    if method == 'interp':
+        x = np.linspace(0, 1, 101)  # 101-point interp (COCO)
+        ap = np.trapz(np.interp(x, mrec, mpre), x)  # integrate
+    else:  # 'continuous'
+        i = np.where(mrec[1:] != mrec[:-1])[0]  # points where x axis (recall) changes
+        ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])  # area under curve
+
+    return ap, mpre, mrec
+
+
+class ConfusionMatrix:
+    # Updated version of https://github.com/kaanakan/object_detection_confusion_matrix
+    def __init__(self, nc, conf=0.25, iou_thres=0.45):
+        self.matrix = np.zeros((nc + 1, nc + 1))
+        self.nc = nc  # number of classes
+        self.conf = conf
+        self.iou_thres = iou_thres
+
+    def process_batch(self, detections, labels):
+        """
+        Return intersection-over-union (Jaccard index) of boxes.
+        Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
+        Arguments:
+            detections (Array[N, 6]), x1, y1, x2, y2, conf, class
+            labels (Array[M, 5]), class, x1, y1, x2, y2
+        Returns:
+            None, updates confusion matrix accordingly
+        """
+        detections = detections[detections[:, 4] > self.conf]
+        gt_classes = labels[:, 0].int()
+        detection_classes = detections[:, 5].int()
+        iou = general.box_iou(labels[:, 1:], detections[:, :4])
+
+        x = torch.where(iou > self.iou_thres)
+        if x[0].shape[0]:
+            matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy()
+            if x[0].shape[0] > 1:
+                matches = matches[matches[:, 2].argsort()[::-1]]
+                matches = matches[np.unique(matches[:, 1], return_index=True)[1]]
+                matches = matches[matches[:, 2].argsort()[::-1]]
+                matches = matches[np.unique(matches[:, 0], return_index=True)[1]]
+        else:
+            matches = np.zeros((0, 3))
+
+        n = matches.shape[0] > 0
+        m0, m1, _ = matches.transpose().astype(np.int16)
+        for i, gc in enumerate(gt_classes):
+            j = m0 == i
+            if n and sum(j) == 1:
+                self.matrix[gc, detection_classes[m1[j]]] += 1  # correct
+            else:
+                self.matrix[gc, self.nc] += 1  # background FP
+
+        if n:
+            for i, dc in enumerate(detection_classes):
+                if not any(m1 == i):
+                    self.matrix[self.nc, dc] += 1  # background FN
+
+    def matrix(self):
+        return self.matrix
+
+    def plot(self, save_dir='', names=()):
+        try:
+            import seaborn as sn
+
+            array = self.matrix / (self.matrix.sum(0).reshape(1, self.nc + 1) + 1E-6)  # normalize
+            array[array < 0.005] = np.nan  # don't annotate (would appear as 0.00)
+
+            fig = plt.figure(figsize=(12, 9), tight_layout=True)
+            sn.set(font_scale=1.0 if self.nc < 50 else 0.8)  # for label size
+            labels = (0 < len(names) < 99) and len(names) == self.nc  # apply names to ticklabels
+            sn.heatmap(array, annot=self.nc < 30, annot_kws={"size": 8}, cmap='Blues', fmt='.2f', square=True,
+                       xticklabels=names + ['background FN'] if labels else "auto",
+                       yticklabels=names + ['background FP'] if labels else "auto").set_facecolor((1, 1, 1))
+            fig.axes[0].set_xlabel('True')
+            fig.axes[0].set_ylabel('Predicted')
+            fig.savefig(Path(save_dir) / 'confusion_matrix.png', dpi=250)
+        except Exception as e:
+            pass
+
+    def print(self):
+        for i in range(self.nc + 1):
+            print(' '.join(map(str, self.matrix[i])))
+
+
+# Plots ----------------------------------------------------------------------------------------------------------------
+
+def plot_pr_curve(px, py, ap, save_dir='.', names=()):
+    fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)
+    py = np.stack(py, axis=1)
+
+    if 0 < len(names) < 21:  # show mAP in legend if < 10 classes
+        for i, y in enumerate(py.T):
+            ax.plot(px, y, linewidth=1, label=f'{names[i]} %.3f' % ap[i, 0])  # plot(recall, precision)
+    else:
+        ax.plot(px, py, linewidth=1, color='grey')  # plot(recall, precision)
+
+    ax.plot(px, py.mean(1), linewidth=3, color='blue', label='all classes %.3f mAP@0.5' % ap[:, 0].mean())
+    ax.set_xlabel('Recall')
+    ax.set_ylabel('Precision')
+    ax.set_xlim(0, 1)
+    ax.set_ylim(0, 1)
+    plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
+    fig.savefig(Path(save_dir) / 'precision_recall_curve.png', dpi=250)

utils/torch_utils.py

@@ -0,0 +1,242 @@
+# PyTorch utils
+
+import logging
+import math
+import os
+import time
+from contextlib import contextmanager
+from copy import deepcopy
+
+import torch
+import torch.backends.cudnn as cudnn
+import torch.nn as nn
+import torch.nn.functional as F
+import torchvision
+
+logger = logging.getLogger(__name__)
+
+
+@contextmanager
+def torch_distributed_zero_first(local_rank: int):
+    """
+    Decorator to make all processes in distributed training wait for each local_master to do something.
+    """
+    if local_rank not in [-1, 0]:
+        torch.distributed.barrier()
+    yield
+    if local_rank == 0:
+        torch.distributed.barrier()
+
+
+def init_torch_seeds(seed=0):
+    # Speed-reproducibility tradeoff https://pytorch.org/docs/stable/notes/randomness.html
+    torch.manual_seed(seed)
+    if seed == 0:  # slower, more reproducible
+        cudnn.deterministic = True
+        cudnn.benchmark = False
+    else:  # faster, less reproducible
+        cudnn.deterministic = False
+        cudnn.benchmark = True
+
+
+def select_device(device='', batch_size=None):
+    # device = 'cpu' or '0' or '0,1,2,3'
+    cpu_request = device.lower() == 'cpu'
+    if device and not cpu_request:  # if device requested other than 'cpu'
+        os.environ['CUDA_VISIBLE_DEVICES'] = device  # set environment variable
+        assert torch.cuda.is_available(), 'CUDA unavailable, invalid device %s requested' % device  # check availablity
+
+    cuda = False if cpu_request else torch.cuda.is_available()
+    if cuda:
+        c = 1024 ** 2  # bytes to MB
+        ng = torch.cuda.device_count()
+        if ng > 1 and batch_size:  # check that batch_size is compatible with device_count
+            assert batch_size % ng == 0, 'batch-size %g not multiple of GPU count %g' % (batch_size, ng)
+        x = [torch.cuda.get_device_properties(i) for i in range(ng)]
+        s = f'Using torch {torch.__version__} '
+        for i in range(0, ng):
+            if i == 1:
+                s = ' ' * len(s)
+            logger.info("%sCUDA:%g (%s, %dMB)" % (s, i, x[i].name, x[i].total_memory / c))
+    else:
+        logger.info(f'Using torch {torch.__version__} CPU')
+
+    logger.info('')  # skip a line
+    return torch.device('cuda:0' if cuda else 'cpu')
+
+
+def time_synchronized():
+    torch.cuda.synchronize() if torch.cuda.is_available() else None
+    return time.time()
+
+
+def is_parallel(model):
+    return type(model) in (nn.parallel.DataParallel, nn.parallel.DistributedDataParallel)
+
+
+def intersect_dicts(da, db, exclude=()):
+    # Dictionary intersection of matching keys and shapes, omitting 'exclude' keys, using da values
+    return {k: v for k, v in da.items() if k in db and not any(x in k for x in exclude) and v.shape == db[k].shape}
+
+
+def initialize_weights(model):
+    for m in model.modules():
+        t = type(m)
+        if t is nn.Conv2d:
+            pass  # nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
+        elif t is nn.BatchNorm2d:
+            m.eps = 1e-3
+            m.momentum = 0.03
+        elif t in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6]:
+            m.inplace = True
+
+
+def find_modules(model, mclass=nn.Conv2d):
+    # Finds layer indices matching module class 'mclass'
+    return [i for i, m in enumerate(model.module_list) if isinstance(m, mclass)]
+
+
+def sparsity(model):
+    # Return global model sparsity
+    a, b = 0., 0.
+    for p in model.parameters():
+        a += p.numel()
+        b += (p == 0).sum()
+    return b / a
+
+
+def prune(model, amount=0.3):
+    # Prune model to requested global sparsity
+    import torch.nn.utils.prune as prune
+    print('Pruning model... ', end='')
+    for name, m in model.named_modules():
+        if isinstance(m, nn.Conv2d):
+            prune.l1_unstructured(m, name='weight', amount=amount)  # prune
+            prune.remove(m, 'weight')  # make permanent
+    print(' %.3g global sparsity' % sparsity(model))
+
+
+def fuse_conv_and_bn(conv, bn):
+    # Fuse convolution and batchnorm layers https://tehnokv.com/posts/fusing-batchnorm-and-conv/
+    fusedconv = nn.Conv2d(conv.in_channels,
+                          conv.out_channels,
+                          kernel_size=conv.kernel_size,
+                          stride=conv.stride,
+                          padding=conv.padding,
+                          groups=conv.groups,
+                          bias=True).requires_grad_(False).to(conv.weight.device)
+
+    # prepare filters
+    w_conv = conv.weight.clone().view(conv.out_channels, -1)
+    w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var)))
+    fusedconv.weight.copy_(torch.mm(w_bn, w_conv).view(fusedconv.weight.size()))
+
+    # prepare spatial bias
+    b_conv = torch.zeros(conv.weight.size(0), device=conv.weight.device) if conv.bias is None else conv.bias
+    b_bn = bn.bias - bn.weight.mul(bn.running_mean).div(torch.sqrt(bn.running_var + bn.eps))
+    fusedconv.bias.copy_(torch.mm(w_bn, b_conv.reshape(-1, 1)).reshape(-1) + b_bn)
+
+    return fusedconv
+
+
+def model_info(model, verbose=False, img_size=640):
+    # Model information. img_size may be int or list, i.e. img_size=640 or img_size=[640, 320]
+    n_p = sum(x.numel() for x in model.parameters())  # number parameters
+    n_g = sum(x.numel() for x in model.parameters() if x.requires_grad)  # number gradients
+    if verbose:
+        print('%5s %40s %9s %12s %20s %10s %10s' % ('layer', 'name', 'gradient', 'parameters', 'shape', 'mu', 'sigma'))
+        for i, (name, p) in enumerate(model.named_parameters()):
+            name = name.replace('module_list.', '')
+            print('%5g %40s %9s %12g %20s %10.3g %10.3g' %
+                  (i, name, p.requires_grad, p.numel(), list(p.shape), p.mean(), p.std()))
+
+    try:  # FLOPS
+        from thop import profile
+        stride = int(model.stride.max())
+        img = torch.zeros((1, 3, stride, stride), device=next(model.parameters()).device)  # input
+        flops = profile(deepcopy(model), inputs=(img,), verbose=False)[0] / 1E9 * 2  # stride FLOPS
+        img_size = img_size if isinstance(img_size, list) else [img_size, img_size]  # expand if int/float
+        fs = ', %.1f GFLOPS' % (flops * img_size[0] / stride * img_size[1] / stride)  # 640x640 FLOPS
+    except (ImportError, Exception):
+        fs = ''
+
+    logger.info(f"Model Summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}")
+
+
+def load_classifier(name='resnet101', n=2):
+    # Loads a pretrained model reshaped to n-class output
+    model = torchvision.models.__dict__[name](pretrained=True)
+
+    # ResNet model properties
+    # input_size = [3, 224, 224]
+    # input_space = 'RGB'
+    # input_range = [0, 1]
+    # mean = [0.485, 0.456, 0.406]
+    # std = [0.229, 0.224, 0.225]
+
+    # Reshape output to n classes
+    filters = model.fc.weight.shape[1]
+    model.fc.bias = nn.Parameter(torch.zeros(n), requires_grad=True)
+    model.fc.weight = nn.Parameter(torch.zeros(n, filters), requires_grad=True)
+    model.fc.out_features = n
+    return model
+
+
+def scale_img(img, ratio=1.0, same_shape=False):  # img(16,3,256,416), r=ratio
+    # scales img(bs,3,y,x) by ratio
+    if ratio == 1.0:
+        return img
+    else:
+        h, w = img.shape[2:]
+        s = (int(h * ratio), int(w * ratio))  # new size
+        img = F.interpolate(img, size=s, mode='bilinear', align_corners=False)  # resize
+        if not same_shape:  # pad/crop img
+            gs = 32  # (pixels) grid size
+            h, w = [math.ceil(x * ratio / gs) * gs for x in (h, w)]
+        return F.pad(img, [0, w - s[1], 0, h - s[0]], value=0.447)  # value = imagenet mean
+
+
+def copy_attr(a, b, include=(), exclude=()):
+    # Copy attributes from b to a, options to only include [...] and to exclude [...]
+    for k, v in b.__dict__.items():
+        if (len(include) and k not in include) or k.startswith('_') or k in exclude:
+            continue
+        else:
+            setattr(a, k, v)
+
+
+class ModelEMA:
+    """ Model Exponential Moving Average from https://github.com/rwightman/pytorch-image-models
+    Keep a moving average of everything in the model state_dict (parameters and buffers).
+    This is intended to allow functionality like
+    https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage
+    A smoothed version of the weights is necessary for some training schemes to perform well.
+    This class is sensitive where it is initialized in the sequence of model init,
+    GPU assignment and distributed training wrappers.
+    """
+
+    def __init__(self, model, decay=0.9999, updates=0):
+        # Create EMA
+        self.ema = deepcopy(model.module if is_parallel(model) else model).eval()  # FP32 EMA
+        # if next(model.parameters()).device.type != 'cpu':
+        #     self.ema.half()  # FP16 EMA
+        self.updates = updates  # number of EMA updates
+        self.decay = lambda x: decay * (1 - math.exp(-x / 2000))  # decay exponential ramp (to help early epochs)
+        for p in self.ema.parameters():
+            p.requires_grad_(False)
+
+    def update(self, model):
+        # Update EMA parameters
+        with torch.no_grad():
+            self.updates += 1
+            d = self.decay(self.updates)
+
+            msd = model.module.state_dict() if is_parallel(model) else model.state_dict()  # model state_dict
+            for k, v in self.ema.state_dict().items():
+                if v.dtype.is_floating_point:
+                    v *= d
+                    v += (1. - d) * msd[k].detach()
+
+    def update_attr(self, model, include=(), exclude=('process_group', 'reducer')):
+        # Update EMA attributes
+        copy_attr(self.ema, model, include, exclude)