search

AlvinYH / Faster-VoxelPose

Official implementation of Faster VoxelPose: Real-time 3D Human Pose Estimation by Orthographic Projection
MIT License
156 stars 18 forks source link

Number of cameras #36

Open CuiSQ opened 1 year ago

CuiSQ commented 1 year ago

Hello! How many cameras did you use in the context of the article?

AlvinYH commented 11 months ago

Hi, @CuiSQ. In our setup, we utilize 5 cameras for the Panoptic and Shelf dataset, while opting for 3 cameras specifically for the Campus dataset.