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

关于实验性能请教 #39

Open
gzshan opened this issue Nov 19, 2019 · 1 comment
Open

关于实验性能请教 #39

gzshan opened this issue Nov 19, 2019 · 1 comment

Comments

@gzshan
Copy link

gzshan commented Nov 19, 2019

你好,
非常感谢实现了这么好的代码,最近基本将整个源码整体读了一遍。
在我复现的结果中,各项测试指标与你所提供的值基本吻合,但是在与其他方法对比的过程中,发现也有很多类似用到batch hard triplet loss的方法,但是发现此方法得到的结果超过了大多数的同类方法,因此对这一点有点疑惑。
按我理解,这个方法不同的可能在于last_conv_stride,在数据增强上只做了水平翻转,网络模型也很简洁,只提取了简单的全局特征,想探讨一下您认为是什么因素导致性能优于其他同类方法呢,是学习率设置、epoch等等的这些训练技巧吗?

@huanghoujing
Copy link
Owner

您好,感谢关注!
性能提升来自于

  • stride = 1
  • 增加训练迭代次数,300 epochs
  • backbone后面不引入全连接层

另外,batch内P(多少人)、K(每个人多少张图片)两个参数也会影响性能

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