Open-H-Embodiment
A community-driven dataset initiative building the shared foundation to train and evaluate generalist vision-language-action models for healthcare robotics.
Open healthcare and life sciences
A community-led foundation for shared datasets and open models across Healthcare and Life Sciences.
A shared foundation
Open-H brings academic groups, clinicians, and industry teams together to build research resources that are useful beyond any single institution or platform.
Our current initiatives span embodied robotic data, endoluminal robotics, and raw ultrasound channel data.
Current work
Each effort creates a practical route for the community to contribute data, methods, and domain expertise.
A community-driven dataset initiative building the shared foundation to train and evaluate generalist vision-language-action models for healthcare robotics.
An emerging effort focused on endoluminal robotics data and the shared infrastructure needed to advance minimally invasive robotic systems.
Open private repositoryGitHub access is currently required.
An open effort to create high-quality raw ultrasound channel data and foundation models for reconstruction, flow, quantitative imaging, motion estimation, and interpretation.
View GitHub repositoryParticipate
Open-H grows through contributed datasets, technical collaboration, and expert clinical annotation.
Academic and industry groups
Start with the GitHub home for the initiative that matches your data. Repository documentation contains the current scope, formats, and contribution process.
Clinicians
Register for the Open-H annotation platform to review healthcare images and video, contribute expert labels, and take part in active annotation projects.
Register to annotate
Open by design
Progress in healthcare and life sciences requires broad, representative data and shared technical foundations. Open-H is building them with the community.