MM-OR is a pioneering large-scale multimodal dataset designed to advance semantic understanding of operating room environments, enabling more effective surgical assistance systems and improved patient safety.

• First comprehensive dataset capturing the complex interactions among medical staff, tools, and equipment in high-intensity surgical settings
• Addresses critical limitations of existing datasets in scale, realism, and multimodal representation
• Enables development of AI systems that can understand the spatiotemporal dynamics of surgical environments
• Creates foundation for next-generation intelligent assistance tools for surgical teams

This research transforms medical AI capabilities by providing the rich, realistic data needed to build systems that can understand and adapt to the complex, high-stakes environment of modern operating rooms.

MM-OR: A Large Multimodal Operating Room Dataset for Semantic Understanding of High-Intensity Surgical Environments