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

你好,请问detect时候,为啥使用百度盘中的权重voc_77.8.pth,为啥会出现很多框? #26

Open
liukangji opened this issue Apr 3, 2022 · 10 comments

Comments

@liukangji
Copy link

image

@liukangji liukangji changed the title 请问detect时候,使用百度盘中的权重voc_77.8.pth,为啥会出现很多框? 你好,请问detect时候,为啥使用百度盘中的权重voc_77.8.pth,为啥会出现很多框? Apr 3, 2022
@bamboopu
Copy link

bamboopu commented Apr 8, 2022

这个应该是正常现象,因为MAP算法是根据置信度从高到低排序计算recall precision,置信度低的结果并不会从预测结果中被剔除,因而形如图上0.348 0.308的置信度结果都画出来了,实际应用的时候可以自己手动删去低置信度结果。

@liukangji
Copy link
Author

好的,谢谢。还想请教一下,这份程序并没有写loss相关代码,如果我想加的话,在网上也没有找到类似代码,你有什么建议吗?

@bamboopu
Copy link

bamboopu commented Apr 8, 2022

这个程序是有loss代码的,在$FCOS/model/loss.py里面,如果你是新手入门,建议你对着FCOSDetector的forward()函数走一遍,知道每一个模块在调用什么语句从哪调用

@liukangji
Copy link
Author

谢谢,抱歉,我表达错误了,我想把loss图画出来的那种代码。

@bamboopu
Copy link

bamboopu commented Apr 8, 2022

画loss图的操作比较简单,经典操作是查一下tensorboardx的用法就可以(语句少把每轮的loss add一下就行),新的可视化工具也可以用wandb

@liukangji
Copy link
Author

好的,谢谢

@liukangji
Copy link
Author

liukangji commented Apr 8, 2022

画loss图的操作比较简单,经典操作是查一下tensorboardx的用法就可以(语句少把每轮的loss add一下就行),新的可视化工具也可以用wandb

冒昧的问一下,方便加你联系方式,请教一下吗?

@bamboopu
Copy link

bamboopu commented Apr 8, 2022

抱歉,精力有限暂不添加,请善用搜索工具。

@liukangji
Copy link
Author

好的,麻烦了。

@zhenghao977
Copy link
Owner

zhenghao977 commented Oct 11, 2022 via email

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

3 participants