-
Notifications
You must be signed in to change notification settings - Fork 25
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
增加有更多噪声的数据集 #19
Comments
另外,我觉得,除了添加噪声来构建降噪数据集,还可以构建一些被暴力二极化的图片作为输入,我经常会在一些PDF扫描版教材上见到这种情况。 |
好的,我了解一下 |
你说的没错,不过这种问题必然要通过多模态的模型解决,因为要有zero-shot能力。单靠视觉特征肯定不行。后续我会考虑构建一个多模态模型。 |
这种情况是不是与仓库节目修复出therofore的情况类似,都需要多模态能力 |
是的,本质上DocDiff不理解这个Therefore的含义,仅仅根据视觉特征进行还原。加入Text Prior效果会好很多 |
这个开源项目提供了一些生成噪声的预处理流程。
https://github.com/sparkfish/shabby-pages
他提供了一个pipeline,用于合成各种噪声的图片。
The text was updated successfully, but these errors were encountered: