这里显示车辆父类中没有图像路径这个属性，谁能帮我解答一下？

首先感谢您非常棒的作品！！
我换用其他的路况图片存在有的车辆未检测。想换用yolov3的代替tiny。但发现出现了无法检测。请问是什么原因？

谢谢作者提供的代码和引导，但是我发现有些车辆你漏检了，请问是什么原因呢

运行 train_vehicle_multilabel.py 时，就是训练时，显示根目录不存在，不应该就是放数据集根目录的文件吗？那还是放哪个？

你好，请问车辆朝向检测的输出是什么？在脚本中哪里体现呢？

你好，我看你在多标签分类中用到了B-CNN，在VehicleDC.py中这样写的：
“X = torch.bmm(X, torch.transpose(X, 1, 2)) / (1 ** 2) # Bi-linear CNN”
我看了B-CNN的代码，并没有看到这样的方法。
请问你这样做的根据和目的是什么？
也可能是我看漏掉了，还请作者指点。非常感谢

我想问一下车辆多标签文件目录如何放置，我显示一直找不到文件。

而且添加torch也失败了哎

请问能不能把在您自己数据集上的accuracy、loss计算结果等性能结果贴出来参考一下？感激不尽！

@CaptainEven thanks for sharing the code base can you mention the inference time with and without ur approach

Tranied models-vehicle detection https://pan.baidu.com/s/1HwTCVGTmdqkeLnqnxfNL8Q
Tranied models-vehicle classification https://pan.baidu.com/s/1XmzjvCgOrrVv0NWTt4Fm3g
在运行Vehicle_DC脚本之前，先下载上面的模型文件或者使用自己预先训练好的模型文件，将car_540000.weights（用于检测）放在项目根目录，将epoch_39.pth（用于多标签识别）放在根目录下的checkpoints目录下，即可使用Vehicle_DC运行。
Before running Vehicle_DC,
I cannot download the datasets from the provided links. Is it possible an update, please?
Thank you

作者可以添加一个对视频进行识别的函数吗?谢谢！

只对车辆类型训练的时候,训练精度高而测试精度一直在70%左右呢?

作者好，我在1080和1660ti的机器上跑你的程序，目标检测的速度和识别的速度都达不到你描述的10ms、18ms的速度，请问还有什么技巧吗？希望尽快回复，谢谢

请问在您的测试数据集上准确识别率有多少

Thank you very much for your outstanding contribution. I have a question. What are the requirements of this network for inputting pictures? Why do I input other pictures and report errors? The pictures in your test set will not report errors. The errors are as follows: tile cannot extend outside image. If you can reply, I will be very grateful!

hello,could you please supply requirements.txt file in your project?

Please upload files behind these links:

Tranied models-vehicle detection
Tranied models-vehicle classification

on GitHub so everyone can use them.

Everyone who doesn't own Chinese mobile number cannot download these files.
Why putting project on GitHub than at all than when only Chinese can use it?

Hi Captain Even
I got the following error message while I run vehicleDC.py, I use the yolov3.cfg from pytorch yolo v3, and the weight is download from your network disk.
hope I can get your help. many thanks.
cheers
Edward

DR_model = Car_DC(src_dir=args.src_dir, dst_dir=args.dst_dir)

File "D:/Dev/Vehicle-Car-detection-and-multilabel-classification-master/VehicleDC.py", line 240, in init
self.detector.load_weights(car_det_weights_path)

File "D:\Dev\vehicle\Vehicle-Car-detection-and-multilabel-classification-master\darknet.py", line 440, in load_weights
conv_weights = conv_weights.view_as(conv.weight.data)

RuntimeError: invalid argument 2: size '[512 x 256 x 3 x 3]' is invalid for input with 209204 elements at ..\aten\src\TH\THStorage.cpp:84

The output topK's K is fixed now.

Do you think training a classifier to predict the value of K for every input is a good solution?

Thank you very much.

@CaptainEven ， 谢谢你的分享，我现在有个疑问，最后一层FC出来的结果为什么没有归一化到（0，1）之间。你能帮忙解答一下吗?

Hi!
First of all, thanks for sharing these codes, especially the trained models. I've tested the detection part and the results are astonishingly good given the fact that yolov3-tiny has such a simple structure. So I'm really curious about your training approach.
1.What dataset did you use for detection? I've tried it on coco and voc but the results were way worse than yours.
2.Is there any pre-trained weights for yolov3-tiny? or you just trained from scratch?
Hope you can help me!

Hi,
Can you share you trainig set? Big thanks!

作者目前的模型可以识别车身颜色、朝向和车辆类型3种属性，请问能否自己修改代码，新增自己需要的属性？比如车辆的品牌识别，这个网络能否完成车辆品牌这种细粒度分类任务呢？

您好，请问下这个要怎么设置用CUDA啊，它一直提示尝试用CUDA，但是实际是CPU

请问作者是怎么制作数据集的呢，主要想知道如何准备数据集的标签，我们现在也在做车辆检测和多属性识别

inp_dim = 768这个是怎么计算出来的？
是根据模型来算的吗？

我對車輛標識檢測感興趣，我認為，YOLO是最好的方式。
但我是YOLO的新手，我想幫助你。
您是否有車輛標識檢測經驗？
由於我對培訓自己的數據的知識和計算機的性能非常差。
對不起我的可憐因為它不是我的母語。

Hi
thanks for share your nice project.
but i can not download your model from Bidu website, its chines i think. please if possible for you , send me for my email adress
[email protected]

thanks again

将RepNet-MDNet-VehicleReID项目中的模型下载下来，用在此项目中，出现报错。
=> device: cuda:0
=> ./car_540000.weights loaded.
=> car detection model initiated.
Traceback (most recent call last):
File "VehicleDC.py", line 385, in
DR_model = Car_DC(src_dir=args.src_dir, dst_dir=args.dst_dir)
File "VehicleDC.py", line 246, in init
model_path=local_model_path)
File "VehicleDC.py", line 121, in init
self.net.load_state_dict(torch.load(model_path))
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 839, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for Cls_Net:
Missing key(s) in state_dict: "features.0.weight", "features.1.weight", "features.1.bias", "features.1.running_mean", "features.1.running_var", "features.4.0.conv1.weight", "features.4.0.bn1.weight", "features.4.0.bn1.bias", "features.4.0.bn1.running_mean", "features.4.0.bn1.running_var", "features.4.0.conv2.weight", "features.4.0.bn2.weight", "features.4.0.bn2.bias", "features.4.0.bn2.running_mean", "features.4.0.bn2.running_var", "features.4.1.conv1.weight", "features.4.1.bn1.weight", "features.4.1.bn1.bias", "features.4.1.bn1.running_mean", "features.4.1.bn1.running_var", "features.4.1.conv2.weight", "features.4.1.bn2.weight", "features.4.1.bn2.bias", "features.4.1.bn2.running_mean", "features.4.1.bn2.running_var", "features.5.0.conv1.weight", "features.5.0.bn1.weight", "features.5.0.bn1.bias", "features.5.0.bn1.running_mean", "features.5.0.bn1.running_var", "features.5.0.conv2.weight", "features.5.0.bn2.weight", "features.5.0.bn2.bias", "features.5.0.bn2.running_mean", "features.5.0.bn2.running_var", "features.5.0.downsample.0.weight", "features.5.0.downsample.1.weight", "features.5.0.downsample.1.bias", "features.5.0.downsample.1.running_mean", "features.5.0.downsample.1.running_var", "features.5.1.conv1.weight", "features.5.1.bn1.weight", "features.5.1.bn1.bias", "features.5.1.bn1.running_mean", "features.5.1.bn1.running_var", "features.5.1.conv2.weight", "features.5.1.bn2.weight", "features.5.1.bn2.bias", "features.5.1.bn2.running_mean", "features.5.1.bn2.running_var", "features.6.0.conv1.weight", "features.6.0.bn1.weight", "features.6.0.bn1.bias", "features.6.0.bn1.running_mean", "features.6.0.bn1.running_var", "features.6.0.conv2.weight", "features.6.0.bn2.weight", "features.6.0.bn2.bias", "features.6.0.bn2.running_mean", "features.6.0.bn2.running_var", "features.6.0.downsample.0.weight", "features.6.0.downsample.1.weight", "features.6.0.downsample.1.bias", "features.6.0.downsample.1.running_mean", "features.6.0.downsample.1.running_var", "features.6.1.conv1.weight", "features.6.1.bn1.weight", "features.6.1.bn1.bias", "features.6.1.bn1.running_mean", "features.6.1.bn1.running_var", "features.6.1.conv2.weight", "features.6.1.bn2.weight", "features.6.1.bn2.bias", "features.6.1.bn2.running_mean", "features.6.1.bn2.running_var", "features.7.0.conv1.weight", "features.7.0.bn1.weight", "features.7.0.bn1.bias", "features.7.0.bn1.running_mean", "features.7.0.bn1.running_var", "features.7.0.conv2.weight", "features.7.0.bn2.weight", "features.7.0.bn2.bias", "features.7.0.bn2.running_mean", "features.7.0.bn2.running_var", "features.7.0.downsample.0.weight", "features.7.0.downsample.1.weight", "features.7.0.downsample.1.bias", "features.7.0.downsample.1.running_mean", "features.7.0.downsample.1.running_var", "features.7.1.conv1.weight", "features.7.1.bn1.weight", "features.7.1.bn1.bias", "features.7.1.bn1.running_mean", "features.7.1.bn1.running_var", "features.7.1.conv2.weight", "features.7.1.bn2.weight", "features.7.1.bn2.bias", "features.7.1.bn2.running_mean", "features.7.1.bn2.running_var", "fc.weight", "fc.bias".
Unexpected key(s) in state_dict: "conv1_1.weight", "conv1_1.bias", "conv1_3.weight", "conv1_3.bias", "conv1.0.weight", "conv1.0.bias", "conv1.2.weight", "conv1.2.bias", "conv2_1.weight", "conv2_1.bias", "conv2_3.weight", "conv2_3.bias", "conv2.0.weight", "conv2.0.bias", "conv2.2.weight", "conv2.2.bias", "conv3_1.weight", "conv3_1.bias", "conv3_3.weight", "conv3_3.bias", "conv3_5.weight", "conv3_5.bias", "conv3.0.weight", "conv3.0.bias", "conv3.2.weight", "conv3.2.bias", "conv3.4.weight", "conv3.4.bias", "conv4_1_1.weight", "conv4_1_1.bias", "conv4_1_3.weight", "conv4_1_3.bias", "conv4_1_5.weight", "conv4_1_5.bias", "conv4_1.0.weight", "conv4_1.0.bias", "conv4_1.2.weight", "conv4_1.2.bias", "conv4_1.4.weight", "conv4_1.4.bias", "conv4_2_1.weight", "conv4_2_1.bias", "conv4_2_3.weight", "conv4_2_3.bias", "conv4_2_5.weight", "conv4_2_5.bias", "conv4_2.0.weight", "conv4_2.0.bias", "conv4_2.2.weight", "conv4_2.2.bias", "conv4_2.4.weight", "conv4_2.4.bias", "conv5_1_1.weight", "conv5_1_1.bias", "conv5_1_3.weight", "conv5_1_3.bias", "conv5_1_5.weight", "conv5_1_5.bias", "conv5_1.0.weight", "conv5_1.0.bias", "conv5_1.2.weight", "conv5_1.2.bias", "conv5_1.4.weight", "conv5_1.4.bias", "conv5_2_1.weight", "conv5_2_1.bias", "conv5_2_3.weight", "conv5_2_3.bias", "conv5_2_5.weight", "conv5_2_5.bias", "conv5_2.0.weight", "conv5_2.0.bias", "conv5_2.2.weight", "conv5_2.2.bias", "conv5_2.4.weight", "conv5_2.4.bias", "FC6_1_1.weight", "FC6_1_1.bias", "FC6_1_4.weight", "FC6_1_4.bias", "FC6_1.0.weight", "FC6_1.0.bias", "FC6_1.3.weight", "FC6_1.3.bias", "FC6_2_1.weight", "FC6_2_1.bias", "FC6_2_4.weight", "FC6_2_4.bias", "FC6_2.0.weight", "FC6_2.0.bias", "FC6_2.3.weight", "FC6_2.3.bias", "FC7_1.weight", "FC7_1.bias", "FC7_2.weight", "FC7_2.bias", "FC_8.weight", "FC_8.bias", "attrib_classifier.weight", "attrib_classifier.bias", "arc_fc_br2.weight", "arc_fc_br3.weight", "shared_layers.0.0.weight", "shared_layers.0.0.bias", "shared_layers.0.2.weight", "shared_layers.0.2.bias", "shared_layers.1.0.weight", "shared_layers.1.0.bias", "shared_layers.1.2.weight", "shared_layers.1.2.bias", "shared_layers.2.0.weight", "shared_layers.2.0.bias", "shared_layers.2.2.weight", "shared_layers.2.2.bias", "shared_layers.2.4.weight", "shared_layers.2.4.bias", "branch_1_feats.0.0.0.weight", "branch_1_feats.0.0.0.bias", "branch_1_feats.0.0.2.weight", "branch_1_feats.0.0.2.bias", "branch_1_feats.0.1.0.weight", "branch_1_feats.0.1.0.bias", "branch_1_feats.0.1.2.weight", "branch_1_feats.0.1.2.bias", "branch_1_feats.0.2.0.weight", "branch_1_feats.0.2.0.bias", "branch_1_feats.0.2.2.weight", "branch_1_feats.0.2.2.bias", "branch_1_feats.0.2.4.weight", "branch_1_feats.0.2.4.bias", "branch_1_feats.1.0.weight", "branch_1_feats.1.0.bias", "branch_1_feats.1.2.weight", "branch_1_feats.1.2.bias", "branch_1_feats.1.4.weight", "branch_1_feats.1.4.bias", "branch_1_feats.2.0.weight", "branch_1_feats.2.0.bias", "branch_1_feats.2.2.weight", "branch_1_feats.2.2.bias", "branch_1_feats.2.4.weight", "branch_1_feats.2.4.bias", "branch_1_fc.0.0.weight", "branch_1_fc.0.0.bias", "branch_1_fc.0.3.weight", "branch_1_fc.0.3.bias", "branch_1_fc.1.weight", "branch_1_fc.1.bias", "branch_1.0.0.0.0.weight", "branch_1.0.0.0.0.bias", "branch_1.0.0.0.2.weight", "branch_1.0.0.0.2.bias", "branch_1.0.0.1.0.weight", "branch_1.0.0.1.0.bias", "branch_1.0.0.1.2.weight", "branch_1.0.0.1.2.bias", "branch_1.0.0.2.0.weight", "branch_1.0.0.2.0.bias", "branch_1.0.0.2.2.weight", "branch_1.0.0.2.2.bias", "branch_1.0.0.2.4.weight", "branch_1.0.0.2.4.bias", "branch_1.0.1.0.weight", "branch_1.0.1.0.bias", "branch_1.0.1.2.weight", "branch_1.0.1.2.bias", "branch_1.0.1.4.weight", "branch_1.0.1.4.bias", "branch_1.0.2.0.weight", "branch_1.0.2.0.bias", "branch_1.0.2.2.weight", "branch_1.0.2.2.bias", "branch_1.0.2.4.weight", "branch_1.0.2.4.bias", "branch_1.1.0.0.weight", "branch_1.1.0.0.bias", "branch_1.1.0.3.weight", "branch_1.1.0.3.bias", "branch_1.1.1.weight", "branch_1.1.1.bias", "branch_2_feats.0.0.0.weight", "branch_2_feats.0.0.0.bias", "branch_2_feats.0.0.2.weight", "branch_2_feats.0.0.2.bias", "branch_2_feats.0.1.0.weight", "branch_2_feats.0.1.0.bias", "branch_2_feats.0.1.2.weight", "branch_2_feats.0.1.2.bias", "branch_2_feats.0.2.0.weight", "branch_2_feats.0.2.0.bias", "branch_2_feats.0.2.2.weight", "branch_2_feats.0.2.2.bias", "branch_2_feats.0.2.4.weight", "branch_2_feats.0.2.4.bias", "branch_2_feats.1.0.weight", "branch_2_feats.1.0.bias", "branch_2_feats.1.2.weight", "branch_2_feats.1.2.bias", "branch_2_feats.1.4.weight", "branch_2_feats.1.4.bias", "branch_2_feats.2.0.weight", "branch_2_feats.2.0.bias", "branch_2_feats.2.2.weight", "branch_2_feats.2.2.bias", "branch_2_feats.2.4.weight", "branch_2_feats.2.4.bias", "branch_2_fc.0.0.weight", "branch_2_fc.0.0.bias", "branch_2_fc.0.3.weight", "branch_2_fc.0.3.bias", "branch_2_fc.1.weight", "branch_2_fc.1.bias", "branch_2.0.0.0.0.weight", "branch_2.0.0.0.0.bias", "branch_2.0.0.0.2.weight", "branch_2.0.0.0.2.bias", "branch_2.0.0.1.0.weight", "branch_2.0.0.1.0.bias", "branch_2.0.0.1.2.weight", "branch_2.0.0.1.2.bias", "branch_2.0.0.2.0.weight", "branch_2.0.0.2.0.bias", "branch_2.0.0.2.2.weight", "branch_2.0.0.2.2.bias", "branch_2.0.0.2.4.weight", "branch_2.0.0.2.4.bias", "branch_2.0.1.0.weight", "branch_2.0.1.0.bias", "branch_2.0.1.2.weight", "branch_2.0.1.2.bias", "branch_2.0.1.4.weight", "branch_2.0.1.4.bias", "branch_2.0.2.0.weight", "branch_2.0.2.0.bias", "branch_2.0.2.2.weight", "branch_2.0.2.2.bias", "branch_2.0.2.4.weight", "branch_2.0.2.4.bias", "branch_2.1.0.0.weight", "branch_2.1.0.0.bias", "branch_2.1.0.3.weight", "branch_2.1.0.3.bias", "branch_2.1.1.weight", "branch_2.1.1.bias".
请问我需要如何调整，才可以运行起来。