We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
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,我用自己的数据集训练,mAP接近98%、每一类精准率都很高,接近99%,召回率60%左右 但是用测试集图片检测时,大部分目标置信度低(0.3-0.5),必须在代码中调低score值才可以检测出这部分目标, anchor box是自己聚类生成的,匹配度大概98% 不知道这种情况如何解决?求指点
The text was updated successfully, but these errors were encountered:
1、map这么高你的数据集是不是差异性不大。 2、计算map并不是要求所有的正样本置信度都很高。 3、看看map的原理,召回率和精确度的该奶奶
Sorry, something went wrong.
概念
No branches or pull requests
你好,我用自己的数据集训练,mAP接近98%、每一类精准率都很高,接近99%,召回率60%左右
但是用测试集图片检测时,大部分目标置信度低(0.3-0.5),必须在代码中调低score值才可以检测出这部分目标,
anchor box是自己聚类生成的,匹配度大概98%
不知道这种情况如何解决?求指点
The text was updated successfully, but these errors were encountered: