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

center loss #8

Open
yida88 opened this issue May 16, 2019 · 2 comments
Open

center loss #8

yida88 opened this issue May 16, 2019 · 2 comments

Comments

@yida88
Copy link

yida88 commented May 16, 2019

请问,可以单独使用center loss不加softmax loss吗,效果怎么样

@MDR-EX1000
Copy link

@yida88 center loss需要估计一个比较准确的具有判别性的center,这一点是由softmax loss的分类监督来实现的,如果已经有了一个比较准确的center作初始化(或者说有了一个比较好的pre-train模型用于特征提取) 那么只使用center loss很有可能可以改进这个模型... 不过一般来说加上softmax loss来保持这个center的判别性并没有什么坏处 除非softmax loss带来的额外开销不可接受

@ghost
Copy link

ghost commented Jul 7, 2020

center loss 在你给的实验训练结果中并没有明显的趋势。不规律的变化,不是下降的。

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

2 participants