VillardX
VillardX
Thanks for your reply. I think the preprocess of my own data is the same as you say in 1. and 2. However, when I train it from scratch or...
Thanks for reply. My nvs dataset actually assumes a strictly rigid scene. So in your opinion, it's better to finetinue VGGT based on camera_head+depth_head's output, instead of point_head's output? Actually,...
> Hello, I saw similar ideas in a recent method called AnySplat. They used a frozen backbone initialized with VGGT weights, jointly trained the Gaussian head, camera head, and depth...
> > Thanks for reply. My nvs dataset actually assumes a strictly rigid scene. So in your opinion, it's better to finetinue VGGT based on camera_head+depth_head's output, instead of point_head's...
好的，感谢回复~