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

Confused about deformation field #52

Open
ricshaw opened this issue Mar 25, 2022 · 0 comments
Open

Confused about deformation field #52

ricshaw opened this issue Mar 25, 2022 · 0 comments

Comments

@ricshaw
Copy link

ricshaw commented Mar 25, 2022

Hi, thanks for the great work!
I'm trying to get a textured mesh from my nerfie capture but I'm confused about how the deformation field works.
I can get a canonical mesh (no warp) by feeding in a voxel grid of points, predicting alphas, and doing marching cubes to get a mesh of vertices and triangles. Now I want to warp this mesh using the deformation field into each of my training images. But when I feed in these mesh vertices as nerf points, the resulting warped points (out['warped_points']) don't align with my training images when I project them into each camera using camera.project(). However, if I feed in a voxel grid and do marching cubes over the warped alphas for each image, then the resulting vertices line up perfectly with my training images.
I don't understand why it works with the voxel grid, but doesn't work with the mesh vertices. It seems like deforming the points is not the same as getting the deformed alpha and doing marching cubes on that.

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

1 participant