cvlab-columbia / zero123

Zero-1-to-3: Zero-shot One Image to 3D Object (ICCV 2023)
https://zero123.cs.columbia.edu/
MIT License
2.67k stars 192 forks source link

test/benchmark on non-object data #123

Open segalinc opened 7 months ago

segalinc commented 7 months ago

Hello, thanks for your contribution I am curious, have you tested the model on trying to generate consistent multiview of a room for instance and not only on an object? Thank you