Multi-Person Absolute 3D Human Pose Estimation with Weak Depth Supervision

8 Apr 2020  ·  Marton Veges, Andras Lorincz ·

In 3D human pose estimation one of the biggest problems is the lack of large, diverse datasets. This is especially true for multi-person 3D pose estimation, where, to our knowledge, there are only machine generated annotations available for training. To mitigate this issue, we introduce a network that can be trained with additional RGB-D images in a weakly supervised fashion. Due to the existence of cheap sensors, videos with depth maps are widely available, and our method can exploit a large, unannotated dataset. Our algorithm is a monocular, multi-person, absolute pose estimator. We evaluate the algorithm on several benchmarks, showing a consistent improvement in error rates. Also, our model achieves state-of-the-art results on the MuPoTS-3D dataset by a considerable margin.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
3D Multi-Person Pose Estimation (absolute) MuPoTS-3D WDSPose 3DPCK 37.3 # 7
3D Multi-Person Pose Estimation (root-relative) MuPoTS-3D WDSPose 3DPCK 82.7 # 7

Methods


No methods listed for this paper. Add relevant methods here