cvlab-columbia / zero123

Zero-1-to-3: Zero-shot One Image to 3D Object (ICCV 2023)

Home Page:https://zero123.cs.columbia.edu/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

offline rendering or online rendering

bio-mlhui opened this issue · comments

It seems that the rendered images are generated offline first and used later for training. Is it feasible to incorporate rendering to dataloader and generate rendered images online for training? Will it make training too slow? Since the torch dataloader will fetch new batch data before each iteration ends, is that practical?