search

cvlab-columbia / zero123

Zero-1-to-3: Zero-shot One Image to 3D Object (ICCV 2023)
https://zero123.cs.columbia.edu/
MIT License
2.59k stars 188 forks source link

Incorrect result. #100

Open bootsmakes opened 8 months ago

bootsmakes commented 8 months ago

Hello, I tested zero123-xl on an image with multiple objects and the results(mainly positions) are incorrect, does it only work for images with single object? inputoutput