Hello! While reading the paper associated with the code, I noticed that in the paper (page 5 formula 5 & 6), the L1 is the "nearest neighbor classification loss" while L2 is the "global classification loss", and total loss L = lambda L1 + L2. Whereas in the code line 123-125, it seems to me the losses are a bit mixed up.
I'm wondering why the nearest neighbor classification loss is calculated with ytest and pids, which looks like the global loss.
Thank you for the help!
Hello! While reading the paper associated with the code, I noticed that in the paper (page 5 formula 5 & 6), the L1 is the "nearest neighbor classification loss" while L2 is the "global classification loss", and total loss L = lambda L1 + L2. Whereas in the code line 123-125, it seems to me the losses are a bit mixed up.
123 loss1 = criterion(ytest, pids.view(-1)) 124 loss2 = criterion(cls_scores, labels_test.view(-1)) 125 loss = loss1 + 0.5 * loss2
I'm wondering why the nearest neighbor classification loss is calculated with ytest and pids, which looks like the global loss. Thank you for the help!