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 results with holes #7

Open
lanchen2019 opened this issue Aug 14, 2024 · 3 comments
Open

Question about results with holes #7

lanchen2019 opened this issue Aug 14, 2024 · 3 comments

Comments

@lanchen2019
Copy link

Why there are so many holes in the front but it looks beautiful from the back view?
Front
image
Back
image
I was wondering if I set some parameters wrong.
But changing mc_level could not solver this problem. I tried to set it to 6/7/8, but all of them have holes.

@buaacyw
Copy link
Owner

buaacyw commented Aug 14, 2024

That's just a classical failure case.

@Zeina-T
Copy link

Zeina-T commented Sep 5, 2024

I also faced this issue with many different models. V1 is still more reliable 😞

@hitcslj
Copy link

hitcslj commented Oct 15, 2024

I believe this primarily stems from the sequence order of MeshAnythingV2. In MeshLab, the outward orientation of an object is along the z-axis, which results in the front view facets being at the end of the sequence. For the GPT model, as we move towards later tokens, its control ability weakens and the confidence level decreases. Consequently, strange tokens may be sampled, leading to the phenomenon of damaged facets.

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

4 participants