BATCH: Behavior Alignment for Cross-Embodiment Learning from Humans
cs.RO, cs.AI, cs.CV
Submitted: 2026-11-06
Updated: 2026-11-06
Comments: Project website: https://batch-embodiment.github.io/
Project page: https://batch-embodiment.github.io
License: http://creativecommons.org/licenses/by/4.0/
The gist: Human demonstrations provide a scalable source of diverse manipulation data, but leveraging them for robot learning remains challenging due to differences in embodiment, viewpoint, action spaces…
Abstract
Human demonstrations provide a scalable source of diverse manipulation data, but leveraging them for robot learning remains challenging due to differences in embodiment, viewpoint, action spaces, and temporal structure. We introduce BATCH, a cross-embodiment learning framework that explicitly aligns human and robot behaviors. BATCH learns self-supervised temporal representations with a cycle-consistency objective to establish cross-embodiment correspondences without manual annotation. We further introduce two complementary training objectives that leverage these correspondences within an autoregressive videoaction diffusion model: behavior-aligned co-training and human-to-robot style transfer. On real-world manipulation tasks, BATCH improves generalization to unseen task configurations compared with training on unaligned data and prior cross-embodiment baselines, with additional gains from the style-transfer objective. Our results demonstrate that explicit behavior alignment enables more effective transfer from human demonstrations to robot policies.
Sources
- GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation
- Project Aria: A New Tool for Egocentric Multi-Modal AI Research
- Masquerade: Learning from In-the-wild Human Videos using Data-Editing
- Phantom: Training Robots Without Robots Using Only Human Videos
- Causal World Modeling for Robot Control
- ImMimic: Cross-Domain Imitation from Human Videos via Mapping and Interpolation
- EgoVerse: An Egocentric Human Dataset for Robot Learning from Around the World
- Humanoid Policy ~ Human Policy
- DINOv3
- DexWild: Dexterous Human Interactions for In-the-Wild Robot Policies
- Wan: Open and Advanced Large-Scale Video Generative Models
- Ego2Robot: Scalable Robot Data Synthesis from Egocentric Human Data
- Data Analogies Enable Efficient Cross-Embodiment Transfer
- World Action Models are Zero-shot Policies
- MotionTrans: Human VR Data Enable Motion-Level Learning for Robotic Manipulation Policies
- Native Video-Action Pretraining for Generalizable Robot Control
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation