HALOMI: Learning Humanoid Loco-Manipulation with Active Perception from Human Demonstrations
A scalable framework that learns robust humanoid loco-manipulation from egocentric human demonstrations, combining active perception with precise head–hand trajectory tracking.
BioAgent: A Skill-Orchestrated Robot Agent for Biological Experiments
A skill-orchestrated robot agent for long-horizon biological experiments, integrating task planning, skill matching, execution, monitoring, and recovery into one workflow.
FSAG: Enhancing Human-to-Dexterous-Hand Finger-Specific Affordance Grounding via Diffusion Models
A data-efficient framework that bypasses robot grasp-data collection by exploiting the rich, object-centric semantic priors latent in pretrained generative diffusion models.
SAGE: Scene Graph-Aware Guidance and Execution for Long-Horizon Manipulation Tasks
A novel framework for scene graph-aware guidance and execution, targeting reliable long-horizon manipulation.
An unsupervised multi-view stereoscopic system based on Gaussian Splatting — Low-Light Gaussian Splatting — for reconstruction in near-total darkness.
Meta-Scoop: A Coarse-to-Precise Policy Learning Framework for Precision Scooping Across Task Variations
A coarse-to-precise framework enabling precise scooping operations across diverse task variations.
Triplet2Track: A Hierarchical System with Object-Centric Representations for Reliable Long-Horizon Manipulation
The Triplet-to-Track System (TTS): a closed-loop long-horizon imitation-learning system that uses human videos to reduce reliance on robot-collected data.