EgoPoser: Robust Real-Time Egocentric Pose Estimation from Sparse and Intermittent Observations Everywhere

TL;DR

EgoPoser is a real-time egocentric full-body pose estimation method that works with sparse, intermittent observations, generalizes across users, and outperforms existing approaches in speed and accuracy.

Contribution

It introduces a novel global motion decomposition, a SlowFast module for longer motion capture, and robust modeling from limited headset view data, addressing key limitations of prior methods.

Findings

Outperforms state-of-the-art methods qualitatively and quantitatively

Maintains over 600fps inference speed

Generalizes across different body shapes and environments

Abstract

Full-body egocentric pose estimation from head and hand poses alone has become an active area of research to power articulate avatar representations on headset-based platforms. However, existing methods over-rely on the indoor motion-capture spaces in which datasets were recorded, while simultaneously assuming continuous joint motion capture and uniform body dimensions. We propose EgoPoser to overcome these limitations with four main contributions. 1) EgoPoser robustly models body pose from intermittent hand position and orientation tracking only when inside a headset's field of view. 2) We rethink input representations for headset-based ego-pose estimation and introduce a novel global motion decomposition method that predicts full-body pose independent of global positions. 3) We enhance pose estimation by capturing longer motion time series through an efficient SlowFast module design that maintains computational efficiency. 4) EgoPoser generalizes across various body shapes for different users. We experimentally evaluate our method and show that it outperforms state-of-the-art methods both qualitatively and quantitatively while maintaining a high inference speed of over 600fps. EgoPoser establishes a robust baseline for future work where full-body pose estimation no longer needs to rely on outside-in capture and can scale to large-scale and unseen environments.