LSong

@knazeri Thanks for your kind and quick reply! This work and the results are really cool! It is true that in CelebA the images are all frontal human faces centered...

@xskyz I think maybe you should refer to these lines: https://github.com/knazeri/edge-connect/blob/826f2b8d29d9dfcdbcc501bff36c0cd23b2cb30a/src/dataset.py#L102 Or you could open a new issue, since your problem is different from this issue.

@knazeri Thanks for your helpful comments! While you kindly help me, I have tested more images with your released code and model (thanks again for this!). Now I have some...

@BravoLu For Market, it does not take too much time. If your dataset is too large, you can randomly split it into few chunks and then running re_ranking on these...

Hi @BravoLu , I really appreciate it! I am sorry that the paper has not been accepted yet. You may cite the arXiv version if applicable. In fact, we tried...

4GPU, num_instances=8. Other params remained the same.

Sorry, I don't know how to solve this. For my experience, the BN layer affects the results a lot. So using 2 GPU makes a big difference since the BN...

@anilsh open it with `torch.load()`, no need to use `tar xf`. sorry for the confusing suffix

@mittlin Which file are you dealing with?

@YMing-Li Could you try loading with PyTorch 0.3 / Python 2?