Petr Masopust

It is part of casia dataset but you can create your own. It is in form of: ``` 0000463/043.jpg 0000463 0004971/090.jpg 0004971 0663469/111.jpg 0663469 ... ``` Each line is relative...

Find all faces in image and run arcface on each of them. See my faceserver: https://github.com/ehp/faceserver

Hello, for testing call evaluate method from csv_eval.py. For detection look at my detector script: https://github.com/rainofmine/Face_Attention_Network/pull/8/files#diff-b59bb88a6684acf7d00c6dc58b3a41cc

1. try your model with my fork https://github.com/ehp/Face_Attention_Network (see https://github.com/rainofmine/Face_Attention_Network/pull/8 what is different) 2. post full stacktrace, not only error message

Hello, one tensor is in gpu memory and second in ram (for cpu). Fix is easy - append .cpu() to tensor on line 166 and maybe also to stack on...

I dont understand your question - something is computed on cpu but the main training is done on gpu. Just look at the code.

You can detect more types. E.g. men and women faces so set ID 0 for men faces and ID 1 for women faces. Of course you need proper dataset. For...

Second option - see https://github.com/rainofmine/Face_Attention_Network/blob/68393da155da02d365e50e4118ca428eb9d24eb7/dataloader.py#L50

Hello, there is short description for all parameters directly in the detector.py. model.pt is file with your pretrained model, image.jpg is your input image with people, image_faces_rectangles.jpg is written during...

Models from this and my repository are not compatible.