Flowing from Reasoning to Motion: Learning 3D Hand Trajectory Prediction from Egocentric Human Interaction Videos

TL;DR

This paper introduces EgoMAN, a large-scale egocentric dataset and a reasoning-to-motion framework for 3D hand trajectory prediction that integrates semantic reasoning with motion generation.

Contribution

The work presents the EgoMAN dataset and a novel reasoning-to-motion model that links vision-language reasoning with 3D hand trajectory prediction.

Findings

EgoMAN dataset contains 219K 6DoF trajectories and 3M QA pairs.

EgoMAN model achieves accurate, stage-aware 3D hand trajectories.

Model generalizes well across real-world scenes.

Abstract

Prior works on 3D hand trajectory prediction are constrained by datasets that decouple motion from semantic supervision and by models that weakly link reasoning and action. To address these, we first present the EgoMAN dataset, a large-scale egocentric dataset for interaction stage-aware 3D hand trajectory prediction with 219K 6DoF trajectories and 3M structured QA pairs for semantic, spatial, and motion reasoning. We then introduce the EgoMAN model, a reasoning-to-motion framework that links vision-language reasoning and motion generation via a trajectory-token interface. Trained progressively to align reasoning with motion dynamics, our approach yields accurate and stage-aware trajectories with generalization across real-world scenes.