Contact and Human Dynamics from Monocular Video

TL;DR

This paper introduces a physics-based approach to improve the realism of 3D human motion estimation from monocular video by ensuring physical plausibility through contact prediction and trajectory optimization.

Contribution

It presents a novel contact prediction network trained without manual labels and a physics-based trajectory optimization to produce more realistic human motions from video.

Findings

Significantly more realistic motions than purely kinematic methods.

Improved quantitative measures of kinematic and dynamic plausibility.

Effective on complex contact motions like dancing and sports.

Abstract

Existing deep models predict 2D and 3D kinematic poses from video that are approximately accurate, but contain visible errors that violate physical constraints, such as feet penetrating the ground and bodies leaning at extreme angles. In this paper, we present a physics-based method for inferring 3D human motion from video sequences that takes initial 2D and 3D pose estimates as input. We first estimate ground contact timings with a novel prediction network which is trained without hand-labeled data. A physics-based trajectory optimization then solves for a physically-plausible motion, based on the inputs. We show this process produces motions that are significantly more realistic than those from purely kinematic methods, substantially improving quantitative measures of both kinematic and dynamic plausibility. We demonstrate our method on character animation and pose estimation tasks on dynamic motions of dancing and sports with complex contact patterns.