Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning

About

Parametric human models capture global pose but cannot represent the non-rigid surface dynamics of clothing and soft tissue. Generic scene flow estimates dense motion but breaks down on articulated bodies, where pixel-level supervision is also intractable to acquire. We introduce H-Flow, a dense human scene flow that captures both skeletal kinematics and surface deformation. A unified multi-head transformer estimates flow from monocular video, jointly predicting pose and depth as companion outputs. The challenge lies in the lack of supervision. In place of unattainable labels, we anchor the network in the physics of human motion, encoding geometric, structural, and biomechanical priors as cross-modal training objectives. We further introduce DynAct4D, a high-fidelity synthetic benchmark providing dense flow annotations across diverse subjects, garments, and motions. On standard benchmarks, H-Flow outperforms scene-flow and parametric baselines, and generalizes zero-shot to in-the-wild video. Code, models, and the DynAct4D benchmark will be released upon publication

Zhanbo Huang, Xiaoming Liu, Yu Kong• 2026

Related benchmarks

TaskDatasetResultRank
3D Human Pose Estimation3DPW
PA-MPJPE48.7
137
Pose EstimationFit3D
MPJPE112.8
7
Scene Flow EstimationFit3D (In-Domain)
EPE (mm)24.2
7
Scene Flow EstimationDynAct4D (Out-of-Domain)
EPE (mm)28.2
7
Depth Estimation3DPW
MAE158.6
6
Depth EstimationFit3D
MAE124.5
6
Showing 6 of 6 rows

Other info

Follow for update