search

SUDO-AI-3D / zero123plus

Code repository for Zero123++: a Single Image to Consistent Multi-view Diffusion Base Model.
Apache License 2.0
1.76k stars 123 forks source link

How to get the depth map output? #91

Closed kevinhuangxf closed 3 months ago

kevinhuangxf commented 4 months ago

Thanks for the fantastic work.

Accoding to the latest release I can get the normal map output.

Instead of using depth map prior, I wonder how can I get the depth map output for the generated multiple views?

I've tried transfer normal maps to depth maps but the results are inaccurate.

eliphatfs commented 3 months ago

Depth is an absolute-valued modality unlike normals. Thus, the dynamic range of a natural image is not enough to encode useful information in depths.

eliphatfs commented 3 months ago

Basically, you have to use another model upon the multi-view images for other kinds of prior that can get depths for you.

kevinhuangxf commented 3 months ago

Thank you so much for your prompt response. I'm trying to using depth estimation model based on the multi-view images. I just wonder if there is a "ground truth" depth from the zero123plus model.