Sparkle: A Robust and Versatile Representation for Point Cloud based Human Motion Capture

2026-04-01

AI Summary

This work addresses the challenge of simultaneously preserving geometric detail and ensuring robustness in human motion capture. To this end, we propose SparkleMotion, a hierarchical neural framework that explicitly decouples kinematic structure from geometric shape by integrating skeletal joints and surface anchors into a structured human representation. Our approach introduces a kinematic-geometric factorization mechanism, augmented with geometric continuity constraints and kinematic priors, which significantly enhances robustness to noise, occlusion, and domain shifts while retaining fine-grained geometric details. Extensive experiments demonstrate that SparkleMotion achieves state-of-the-art performance across diverse sensor modalities and complex real-world scenarios, offering both high reconstruction accuracy and strong cross-domain generalization capabilities.

Abstract

Point cloud-based motion capture leverages rich spatial geometry and privacy-preserving sensing, but learning robust representations from noisy, unstructured point clouds remains challenging. Existing approaches face a struggle trade-off between point-based methods (geometrically detailed but noisy) and skeleton-based ones (robust but oversimplified). We address the fundamental challenge: how to construct an effective representation for human motion capture that can balance expressiveness and robustness. In this paper, we propose Sparkle, a structured representation unifying skeletal joints and surface anchors with explicit kinematic-geometric factorization. Our framework, SparkleMotion, learns this representation through hierarchical modules embedding geometric continuity and kinematic constraints. By explicitly disentangling internal kinematic structure from external surface geometry, SparkleMotion achieves state-of-the-art performance not only in accuracy but crucially in robustness and generalization under severe domain shifts, noise, and occlusion. Extensive experiments demonstrate our superiority across diverse sensor types and challenging real-world scenarios.

Problem

• point cloud
• human motion capture
• robust representation
• expressiveness
• geometric noise

Innovation

• point cloud
• motion capture
• structured representation
• kinematic-geometric disentanglement
• robustness