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

数据集 #22

Open
chenyang9799 opened this issue Apr 4, 2020 · 1 comment
Open

数据集 #22

chenyang9799 opened this issue Apr 4, 2020 · 1 comment

Comments

@chenyang9799
Copy link

您说Training set contains 16551 images, and validation set contains 4952 images.
voc2007数据集好像没有这么多图片吧?还是说您用的voc2012?
还有一点是,我下载的官方voc2007数据集,它的/ImageSets/Main/目录下已经有了train.txt和val.txt,您是用的它的现成的还是用代码生成的?

@Zzh-tju
Copy link
Owner

Zzh-tju commented Apr 4, 2020

论文中的VOC实验采取的是VOC 07+12,训练集是VOC2007和VOC2012的全部trainval,评估集是VOC2007 test.
下载数据集后,使用scripts/voc_label.py,就可以自动生成train.txt,共16551张图片。2007_test.txt 4952张图片。

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