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

How to render a picture without eyes? #67

Open
Tianxinhuang opened this issue Oct 18, 2023 · 4 comments
Open

How to render a picture without eyes? #67

Tianxinhuang opened this issue Oct 18, 2023 · 4 comments

Comments

@Tianxinhuang
Copy link

Dear author,

Thank you for the great work. During the reproducing process, I noticed that the rendering result is with eye balls, while the paper results do not have that. I wonder how can we render such a result without eyeball? Thank you very much.

@Inferencer
Copy link

bump, I need this also will install soon but that's why I'm choosing this tool

@Tianxinhuang
Copy link
Author

I find that the eyes can be removed by removing eyes in the mesh models. So, I edit the original average face to achieve this

@Inferencer
Copy link

Inferencer commented Dec 20, 2023

I find that the eyes can be removed by removing eyes in the mesh models. So, I edit the original average face to achieve this

good tip not sure why I was going to remove eyes in post, do you find the outer eye skin alignment is good for single image or does the skin show when overlapped with original footage

also
I can't see the base model to edit, I was expecting to find an obj or similar in either the repo or download links can you point me in the direction of this, perhaps it will appear when I start inference but would be good to know if I'm being blind

@Tianxinhuang
Copy link
Author

Hi, I do this just to enhance the visualization performances after overlapping with the original face image, as this model do not have eye ball model to control the sight direction.
As for the editing, the meshes are actually constructed by vertices and faces. The visualization is controlled by topology of vertices saved in "face32" variable in pipeline.py. If you remove some rows in that variable, corresponding areas will be removed in rendered images. You can open any generated face mesh obj, remove any area with meshlab, and finally get your required topology from the new obj file.

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