Hao Luo

You can write a custom class to build your dataset. You can refer to market.py. The output should be: 

看上去map有点低，rank-1差不多

MODEL.NAME: 'resnet50_ibn_a' PRETRAIN_PATH: Your path

config/defaults.py _C.MODEL.NAME = 'resnet50_ibn_a'

@DTennant Thank you for your great job. It seems this is the best (or second best) performance on Vehicle ReID task. I will introduce it in Readme and never close...

Please refer to https://github.com/michuanhaohao/reid-strong-baseline/issues/101#issuecomment-556838939

You should write an RAP.py by yourself and confirm that self.num_train_pids == 1295. The training set of Market1501 has 751 IDs, so the self.num_train_pids == 751 in Market1501.

You can print the keys of the model.state_dict() and 'resnet50_ibn_1.pth'. The keys may be not same.

Thank you very much! I will keep this issue open. Everyone can discuss it.

Ok, I have not tried these loss functions. Others can discuss them here.