CrossBFM: Distilling a Shared Latent Behavior Space Across Humanoid Embodiments
cs.RO
Submitted: 2026-09-29
Updated: 2026-09-29
Code: https://github.com/Roboparty/UFO
Project page: https://dotandung.github.io/crossbfm
Terminology
Sources
- Retargeting Matters: General Motion Retargeting for Humanoid Motion Tracking
- Learning Successor States and Goal-Dependent Values: A Mathematical Viewpoint
- Multi-Embodiment Locomotion at Scale with extreme Embodiment Randomization
- UniT: Toward a Unified Physical Language for Human-to-Humanoid Policy Learning and World Modeling
- CompliantWBC: Whole-Body Compliance for Heavy Humanoids via Force Latent Estimation and Residual Impedance Targets
- Variational Intrinsic Control
- Classifier-Free Diffusion Guidance
- PHASOR: Phase-Anchored Universal Action Representations for Humanoid Embodiments
- Rapid Embodiment Adaptation for Quadrupedal Locomotion
- BadWAM: When World-Action Models Dream Right but Act Wrong
- Vision-Language-Action Safety: Threats, Challenges, Evaluations, and Mechanisms
- BFM-Zero: A Promptable Behavioral Foundation Model for Humanoid Control Using Unsupervised Reinforcement Learning
- Proximal Policy Optimization Algorithms
- World Action Models: A Survey
- Perceptive Humanoid Parkour: Chaining Dynamic Human Skills via Motion Matching
- Learning a Unified Latent Space for Cross-Embodiment Robot Control
- TWIST2: Scalable, Portable, and Holistic Humanoid Data Collection System
- MOTIF: Learning Action Motifs for Few-shot Cross-Embodiment Transfer
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies
- Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving