GGMotion: Group Graph Dynamics-Kinematics Networks for Human Motion Prediction

TL;DR

GGMotion introduces a novel group graph network that models human motion by capturing complex physical dependencies and priors, significantly improving short-term human motion prediction accuracy.

Contribution

The paper proposes GGMotion, a group graph dynamics-kinematics network with a radial field and equivariant MLPs to better model human motion physics and dependencies.

Findings

Outperforms existing methods on Human3.6M, CMU-Mocap, and 3DPW benchmarks.

Achieves significant improvements in short-term motion prediction accuracy.

Effectively captures multi-scale joint dependencies and physical plausibility.

Abstract

Human motion is a continuous physical process in 3D space, governed by complex dynamic and kinematic constraints. Existing methods typically represent the human pose as an abstract graph structure, neglecting the intrinsic physical dependencies between joints, which increases learning difficulty and makes the model prone to generating unrealistic motions. In this paper, we propose GGMotion, a group graph dynamics-kinematics network that models human topology in groups to better leverage dynamics and kinematics priors. To preserve the geometric equivariance in 3D space, we propose a novel radial field for the graph network that captures more comprehensive spatio-temporal dependencies by aggregating joint features through spatial and temporal edges. Inter-group and intra-group interaction modules are employed to capture the dependencies of joints at different scales. Combined with equivariant multilayer perceptrons (MLP), joint position features are updated in each group through parallelized dynamics-kinematics propagation to improve physical plausibility. Meanwhile, we introduce an auxiliary loss to supervise motion priors during training. Extensive experiments on three standard benchmarks, including Human3.6M, CMU-Mocap, and 3DPW, demonstrate the effectiveness and superiority of our approach, achieving a significant performance margin in short-term motion prediction. The code is available at https://github.com/inkcat520/GGMotion.git.