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

请问40k数据集是指得ImageNet么 #1

Open
kkpssr opened this issue Mar 30, 2022 · 10 comments
Open

请问40k数据集是指得ImageNet么 #1

kkpssr opened this issue Mar 30, 2022 · 10 comments
Labels
question Further information is requested

Comments

@kkpssr
Copy link

kkpssr commented Mar 30, 2022

No description provided.

@kkpssr kkpssr changed the title 请问40k数据集是值得ImageNet么 请问40k数据集是指得ImageNet么 Mar 30, 2022
@Junjue-Wang
Copy link
Owner

不是啊,这是篡改检测任务,ImageNet是分类任务

@kkpssr
Copy link
Author

kkpssr commented Mar 30, 2022 via email

@Junjue-Wang
Copy link
Owner

请问40k数据量的篡改检测集是您这边自己做的么?

---原始邮件--- 发件人: @.> 发送时间: 2022年3月30日(周三) 下午4:56 收件人: @.>; 抄送: @.@.>; 主题: Re: [Junjue-Wang/Rank1-Ali-Tianchi-Real-World-Image-Forgery-Localization-Challenge] 请问40k数据集是指得ImageNet么 (Issue #1) 不是啊,这是篡改检测任务,ImageNet是分类任务 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

这个是上一届的篡改比赛训练数据集,在竞赛论坛里面有公开

1 similar comment
@Junjue-Wang
Copy link
Owner

请问40k数据量的篡改检测集是您这边自己做的么?

---原始邮件--- 发件人: @.> 发送时间: 2022年3月30日(周三) 下午4:56 收件人: @.>; 抄送: @.@.>; 主题: Re: [Junjue-Wang/Rank1-Ali-Tianchi-Real-World-Image-Forgery-Localization-Challenge] 请问40k数据集是指得ImageNet么 (Issue #1) 不是啊,这是篡改检测任务,ImageNet是分类任务 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

这个是上一届的篡改比赛训练数据集,在竞赛论坛里面有公开

@kkpssr
Copy link
Author

kkpssr commented Mar 30, 2022 via email

@Junjue-Wang
Copy link
Owner

目前还没有,主要是觉得我们的trick太多了,模型参数不太具有科研价值

@Junjue-Wang
Copy link
Owner

如想继续研究,请参考我们提出的三个挑战与我们的一些解决思路,希望你能走的更远!

@Junjue-Wang Junjue-Wang added the question Further information is requested label Mar 30, 2022
@kkpssr
Copy link
Author

kkpssr commented Mar 30, 2022 via email

@Junjue-Wang
Copy link
Owner

我的专业不是做篡改检测,意见仅供参考与交流,从此次比赛来看,mvssnet等网络发挥空间不大,很大的原因可能是因为比赛的数据量比较小(泛化性很拉分),但是“挑战2”中的,边缘局部特征增强非常重要的,mvssnet也是基于此,只不过我们的策略是很简单的放大,我相信如果把mvssnet的增强策略放到我们的模型中,应该会有更大的提升。

@kaigelee
Copy link

No description provided.

请问您找到40K的训练数据了吗?能提供个地址吗?感激不尽!
或者邮箱[email protected]

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

3 participants