4 ms·
This is a good question, and the answer is that from a tech side it is surprisingly easier to solve the problem in the reverse direction. As in, making workabl
by stale2002 3y ago
This is a good question, and the answer is that from a tech side it is surprisingly easier to solve the problem in the reverse direction.
As in, making workable 3d models is harder than making video.
And it is easier to make a 3d model by generating a video of the object instead.
Why is that? I don't know. But that's the current state of the industry. 3D model generation is simply harder.
- lvl102 3y agoI am thinking reinforcement on top of Blender would be straight forward with unlimited synthetic data potential. I’ve come across people incorporating SD into rendering workflow so tools are all there.
- euW3EeBe 3y agoProbably also helps that there's way more image/video data to train on than 3D data.