search

ZHU-Zhiyu / NVS_Solver

Source code of paper "NVS-Solver: Video Diffusion Model as Zero-Shot Novel View Synthesizer"
189 stars 1 forks source link

Code question #14

Closed Dumeowmeow closed 5 days ago

Dumeowmeow commented 6 days ago

Thank you for your great work! Did you use the diffusers code in src in your inference process?or directly pip install diffusers?

ZHU-Zhiyu commented 6 days ago

Hi @Dumeowmeow , thanks for your attention on our work. Yes, we install diffusers by pip. Please refer to our installation pipeline. image

Dumeowmeow commented 5 days ago

Thanks for your detailed explanation!

Dumeowmeow commented 5 days ago

Sorry, I have another question. Why step_single twice in svd_interpolate_single_img_traj?

mengyou2 commented 5 days ago

The first one is to compute the backward gradient in eq.14 and the second one is to compute the denoised result.

Dumeowmeow @.***>于2024年7月2日 周二下午5:41写道：

Sorry, I have another question. Why step_single twice in svd_interpolate_single_img_traj?

— Reply to this email directly, view it on GitHub https://github.com/ZHU-Zhiyu/NVS_Solver/issues/14#issuecomment-2202573546, or unsubscribe https://github.com/notifications/unsubscribe-auth/AUI5FS7STV76DA3Y3EGLMFDZKJYT3AVCNFSM6AAAAABKFIOVV6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMBSGU3TGNJUGY . You are receiving this because you are subscribed to this thread.Message ID: @.***>