fengtingl

Not yet. For the best result so far, I use a Proxy-NCA pre-trained network, which achieves a recall@1 rate of around 81%, as the identity discriminator. Then I extend the...

1. I tried two initializations, with the first 1200 identities and a random sampled 1200 identities. The results are not very much distinguishable. 2. I simply change the forward conditional...

Thanks for your reply. I use batch size 16 and learning rate 1e-5 here. Could you provide a pre-trained model? ________________________________ From: Ismail Elezi Sent: Wednesday, December 16, 2020 22:38...