wenqsun/DimensionX

Training Code

Opened this issue · 3 comments

Hello and thank you for this amazing project.

Any plans to release the training code to split the spacial and temporal? That would be incredibly beneficial to the community.

agreed! that would be great

Thanks for your great interest! Now we have released an S-Director, which can be used to generate videos with camera orbit left. More details about our training dataset and code will be released upon paper acceptance.

Can't wait to read the full paper, I'm curious about some details that are in the appendix. Do you have a planned date for full preprint release?