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.

Sign up for GitHub

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

Jump to bottom

关于训练结果的问题 #114

Closed
dexterod opened this issue Nov 6, 2020 · 5 comments
Closed

关于训练结果的问题 #114

dexterod opened this issue Nov 6, 2020 · 5 comments

Comments

@dexterod
Copy link

dexterod commented Nov 6, 2020

您好楼主！

我想问一下，我使用咱们R2CNN程序和训练好的权重文件108000.ckpt，在DOTA的val上跑结果。最终显示mAP只有0.38。然后各种调参数，重新从头训练到50w次了，mAP最多也只到0.48。同样的问题还有RRPN。
有点想不通这精度的差距能差在哪里，会跟机器和配置啥有关吗？还是我有什么地方没有注意到么？

望大哥解惑，谢谢了~~
@yangxue0827
Copy link
Member

yangxue0827 commented Nov 6, 2020 •
edited

建议使用RetinaNet_Tensorflow_Rotation，里面包含了R2CNN-FPN的代码，基于Resnet50的baseline有72+

@dexterod dexterod closed this as completed Nov 6, 2020
@dexterod
Copy link
Author

dexterod commented Nov 6, 2020

建议使用RetinaNet_Tensorflow_Rotation，里面包含了R2CNN-FPN的代码，基于Resnet50的baseline有72+

@dexterod dexterod reopened this Nov 6, 2020
@dexterod
Copy link
Author

dexterod commented Nov 6, 2020

嗯嗯，你的那篇文章我也拜读过了，非常棒，给了我很多启发。但是现在实验得与其他的一些基准算法做比较，我就想用你的这个R2CNN作为标准。所以你测下来60%的精度当时是加了FPN结构是么

@yangxue0827
Copy link
Member

我推荐你的只是一个benchmark，里面基本都是baseline的方法，没有tricks。整个benchmark做过代码结构整理和优化，当前这个repo只有60+存在很多不完善的地方，所以基本已经不维护了。

@dexterod
Copy link
Author

dexterod commented Nov 6, 2020

嗯嗯，好的，那我试试新方法吧，谢谢你~

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@yangxue0827 @dexterod