Person-Attribute-Recognition-MarketDuke
Person-Attribute-Recognition-MarketDuke copied to clipboard
A simple baseline implemented in PyTorch for pedestrian attribute recognition task, evaluating on Market-1501 and DukeMTMC-reID dataset.
Hello, Thank you for this awesome repo. I have a question, why in import_Market1501Attribute.py file, the label order of Train data is changed? Why both train and test label don't...
Duke上训练的模型去识别test_market图像感觉准确率挺低的,不但识别出了bag、handbag,鞋子和裙子的颜色也不对
Is there attribute re-weighting module in the training net?
@hyk1996 thanks ffor opensourcing hte code base , wanted to knw what is the inference time take by the model with your environment per bounding box of a person
run python3 inference.py test_sample/test_market.jpg --dataset market --model resnet50 ---------------------------------------------- RuntimeError: Error(s) in loading state_dict for ResNet50_nFC: Missing key(s) in state_dict: "class_0.classifier.1.weight", "class_0.classifier.1.bias", "class_0.classifier.1.running_mean", "class_0.classifier.1.running_var", "class_0.classifier.4.weight", "class_0.classifier.4.bias", "class_1.classifier.1.weight", "class_1.classifier.1.bias", "class_1.classifier.1.running_mean", "class_1.classifier.1.running_var",...
RuntimeError: size mismatch, m1: [32 x 92160], m2: [2048 x 512]
运行test.py时候提示某些属性没有正负样本警告,所以请问应该在哪增加正负样本?
Hi thanks for your work. im trying to train the model from begin. but found that accuracy does not get better after first epoch of training. here's command i used...
hello: i tried many times such as: python inference.py test_sample/test_market.jpg --dataset market python3 inference.py test_sample/test_market.jpg python inference.py test_sample/test_market.jpg market resnet50 and so on ,but i got the same mistakes the_type...