search

ZHU-Zhiyu / NVS_Solver

Source code of paper "NVS-Solver: Video Diffusion Model as Zero-Shot Novel View Synthesizer"
236 stars 4 forks source link

Difference with SVD #20

Open LeiJiang1 opened 1 month ago

LeiJiang1 commented 1 month ago

Since you are using SVD for generating videos directly, it is more straightforward to compare your method with SVD as well since SVD also supports generating multi-view videos, given a camera pose.

mengyou2 commented 1 month ago

Hi, thanks for your interest in our project. Our project aims to generate novel views of scenes given unknown camera poses. As far as I know, svd cannot input camera pose straightly. did you mean sv3d? But it only generate multiviews of the object, not the whole scene.