Scaling Up Dynamic Human-Scene Interaction Modeling
This work addresses the problem of realistic human-scene interaction synthesis for computer vision and robotics, though it is incremental as it builds on existing motion synthesis methods with a new dataset.
The authors tackled data scarcity in human-scene interaction modeling by introducing the TRUMANS dataset, the most comprehensive motion-captured dataset with over 15 hours of interactions across 100 scenes, and a diffusion-based autoregressive model that generates realistic sequences with strong zero-shot generalizability across multiple 3D scene datasets.
Confronting the challenges of data scarcity and advanced motion synthesis in human-scene interaction modeling, we introduce the TRUMANS dataset alongside a novel HSI motion synthesis method. TRUMANS stands as the most comprehensive motion-captured HSI dataset currently available, encompassing over 15 hours of human interactions across 100 indoor scenes. It intricately captures whole-body human motions and part-level object dynamics, focusing on the realism of contact. This dataset is further scaled up by transforming physical environments into exact virtual models and applying extensive augmentations to appearance and motion for both humans and objects while maintaining interaction fidelity. Utilizing TRUMANS, we devise a diffusion-based autoregressive model that efficiently generates HSI sequences of any length, taking into account both scene context and intended actions. In experiments, our approach shows remarkable zero-shot generalizability on a range of 3D scene datasets (e.g., PROX, Replica, ScanNet, ScanNet++), producing motions that closely mimic original motion-captured sequences, as confirmed by quantitative experiments and human studies.