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

Question about Densification #180

Open
cdfan0627 opened this issue Aug 3, 2024 · 5 comments
Open

Question about Densification #180

cdfan0627 opened this issue Aug 3, 2024 · 5 comments

Comments

@cdfan0627
Copy link

您好,非常感謝您release code,想請問一下我在gaussian_model.py看到你在Densification的時候只考慮了canonical gaussian 的Densification,並沒有考慮deform後的gaussian的Densification,例如下面的連結的code只考慮了canonical gaussian 的scaling,這樣在Densification時是否不太合理呢?
https://github.com/hustvl/4DGaussians/blob/master/scene/gaussian_model.py#L448

@guanjunwu
Copy link
Collaborator

你好,这的确是一种简化的假设,在deform之后的Gaussian作densification按理说是更合理的,但是我们很难将deform之后的Gaussian和canonical Gaussian对应起来(具体可以参考Dynamic Gaussian Mesh,额外训了一个网络,但是这样做会增加计算开销。)

@cdfan0627
Copy link
Author

非常感謝您的解答,想再請問一下你們有試過flow supervision或depth surpervison嗎? 因為我目測試起來發現效果並沒有改善,所以我正在想會不會是densification的原因。

@guanjunwu
Copy link
Collaborator

我们都试过,感觉flow supervision一次好像也只渲染timestamp相邻的2-3张图,所以不是很有用。但是depth在某些单目场景是有成效的。主要还是deformationfield有时候会陷入local minima

@vamWu
Copy link

vamWu commented Aug 15, 2024

非常感謝您的解答,想再請問一下你們有試過flow supervision或depth surpervison嗎? 因為我目測試起來發現效果並沒有改善,所以我正在想會不會是densification的原因。
你好,请问你提到的flow supervision 是指使用optical flow来做监督吗? 我最近也关注到了相关的内容,试图加入一个类似的flowloss来改善reconstruction的效果,但是如您所说似乎提升的很微弱?

@cdfan0627
Copy link
Author

對的,我是用optical flow来做监督,我目前是效果不只沒上升反而有下降

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