Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feature_copy = feature.clone().detach() #1

Open
Walstruzz opened this issue Jul 24, 2024 · 0 comments
Open

feature_copy = feature.clone().detach() #1

Walstruzz opened this issue Jul 24, 2024 · 0 comments

Comments

@Walstruzz
Copy link

感谢您的工作, 我最近在我的工作里复现ECCLoss, 发现在细粒度小数据集上的确提升明显, 不过当我和这里的ECCLoss实现进行对比的时候, 我发现有一些细微的差异, 有一些疑惑的地方, 希望能帮忙解答

ECC/utils.py

Line 29 in 6a21424

feature_copy = feature.clone().detach()

这里detach出来之后, feature_copyfeature_table都没计算梯度, 那feature_intra_loss 是否会回传梯度?

ECC/utils.py

Line 43 in 6a21424

class_table = (class_table - torch.min(class_table)) / (torch.max(class_table) - torch.min(class_table))

这里为什么要做归一化, 我看论文中没有提及, 也可能是我遗漏了什么信息?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant