World-Grounded Human Motion Recovery via Gravity-View Coordinates

TL;DR

This paper introduces a gravity-view coordinate system for monocular human motion recovery, reducing ambiguity and error accumulation, leading to more accurate and realistic world-grounded motion estimation from videos.

Contribution

The novel Gravity-View coordinate system enables direct, gravity-aligned pose estimation per frame, improving accuracy and robustness over autoregressive methods.

Findings

Outperforms state-of-the-art in accuracy and speed

Reduces error accumulation in motion sequences

Produces more realistic world-grounded human motion

Abstract

We present a novel method for recovering world-grounded human motion from monocular video. The main challenge lies in the ambiguity of defining the world coordinate system, which varies between sequences. Previous approaches attempt to alleviate this issue by predicting relative motion in an autoregressive manner, but are prone to accumulating errors. Instead, we propose estimating human poses in a novel Gravity-View (GV) coordinate system, which is defined by the world gravity and the camera view direction. The proposed GV system is naturally gravity-aligned and uniquely defined for each video frame, largely reducing the ambiguity of learning image-pose mapping. The estimated poses can be transformed back to the world coordinate system using camera rotations, forming a global motion sequence. Additionally, the per-frame estimation avoids error accumulation in the autoregressive methods. Experiments on in-the-wild benchmarks demonstrate that our method recovers more realistic motion in both the camera space and world-grounded settings, outperforming state-of-the-art methods in both accuracy and speed. The code is available at https://zju3dv.github.io/gvhmr/.