search

facebookresearch / banmo

BANMo Building Animatable 3D Neural Models from Many Casual Videos
Other
531 stars 59 forks source link

Custom videos for testing #31

Open AbhishekBose opened 2 years ago

AbhishekBose commented 2 years ago

Can I use the pre-trained models on videos containing food items?

gengshan-y commented 2 years ago

The pre-trained PoseNet is for human/quadruped animals. For other categories, initial viewpoint estimation is needed (also see "sensitivity to initial root poses" in the paper supplement).

Also, this codebase uses (slow) test-time optimization and does not support feedforwd inference. (see this issue)