trojans-face-recognizer

trojans-face-recognizer copied to clipboard

about the training codes.

5

Hi, could you share the training codes, like the new loss function and others, thank you so much.

fuxuliu

Doubt regarding ArgNegLoss

1

Hello All, Can someone give me an insight like how does ArgNegLoss weakens the effect of mislabelled dataset. Thanks

random-ai

Have you ever tested it on megaface？

olojuwin

what is the optimal scale and margin value when choose ArcNegFace Loss?

2

billtiger

dingweichao123

In script ./QAN++/test_quality.py

nerddd

Hi, could you share the training codes, like the tricks, thanks a lot.

for-love-forever

Wrong Implementation of the Stochastic Depth ?

1

The authors try to achieve Stochastic Depth by simply set "item.weight.requires_grad" to True/False. However, there are two issues: 1. If a block is drop out during one iteration, all of...

Bear-kai

Could you offer us more information about why ArcNegFace loss is helpful to utilize hard negative mining and weaken the influence of the error labeling? I didn't see any explanation...

meanmee

RuntimeError: you can only change requires_grad flags of leaf variables.

1

"item.weight.requires_grad = True" of the forward function of class BlockA, BlockB and BlockC will cause the above error. I found it is caused by the use of DataParallel of the...

Bear-kai