VR human demos fix persistent failures in point-cloud grasping policies
hcltech-robotics · hf · 2026-11-17
HCL Robotics published a human-in-the-loop failure-mining pipeline on Hugging Face: persistently unsolved grasps of a reinforcement-learning point-cloud policy are turned into virtual-reality human demonstrations.
- Automatically mines grasp cases the policy repeatedly fails on
- VR demonstrations provide corrective signals for policy improvement
- Improves hard-case success rates without reducing generalization to unseen objects
More from Embodied
- Ultrasound plus a smartwatch and a lightweight AI model enables full-body tracking — rschu · 2026-11-17
- DeepMind's Raia Hadsell on transferring skills between robots at RAAIS — nathanbenaich · 2026-11-17
- Ropedia-Xperience-2 dataset: 1,500+ 3D scene scans and 131 hours of egocentric video — liuziwei7 · 2026-11-17
- Niantic's Object Interaction Makes Gaussian Splat Scenes Editable and Physics-Ready — lukas_m_ziegler · 2026-11-17
- A 4.5M-parameter TTS model runs fully offline on an ESP32-S3 — ngxson · 2026-11-17
- NVIDIA open-sources Alpamayo2-Edge, a VLA model for autonomous driving — nvidia · 2026-11-17