NVIDIA and 35 partner organizations have released Open-H-Embodiment, the first large-scale open dataset for healthcare robotics, comprising 778 hours of CC-BY-4.0 surgical, ultrasound, and colonoscopy data. Alongside it, two new open-source models are introduced: GR00T-H, a Vision-Language-Action model trained on ~600 hours of
Table of contents
Introducing Open-H-Embodiment: The first healthcare robotics open dataset, built by a community collaboration 1. Open-H-Embodiment Participants The Dataset 2. GR00T-H: Vision Language Action Model for Surgical Robotics Architectural Design Choices 3. Cosmos-H-Surgical-Simulator Key Capabilities Fine-Tuning Details 4. What is Next: Towards Reasoning For Surgical Robotics 5. Get started today Introducing Open-H-Embodiment: The first healthcare robotics open dataset, built by a community collaboration1. Open-H-Embodiment2. GR00T-H: Vision Language Action Model for Surgical Robotics3. Cosmos-H-Surgical-Simulator4. What is Next: Towards Reasoning For Surgical Robotics5. Get started todaySort: