Building an AI Nurse: Leveraging Synthetic Data for Real-World Impact

At Teton, we're building an AI-powered nurse assistant using cutting-edge computer vision. In this talk, I'll showcase our latest advancements of TEO-1, which is trained on Gefion, the largest supercomputer in the Nordics. We'll explore how we combine procedurally generated 3D data with real-world data to build a robust AI system.

I'll dive into the technical challenges of developing 3D spatial understanding—why it matters and how it helps with occlusion reasoning, multi-camera calibration, and even gait analysis.

Bio: Frederik Warburg is Head of AI at Teton, where he leads groundbreaking research in computer vision and deep learning for healthcare. He completed his PhD at DTU and UC Berkeley, earning multiple awards, including PhD Thesis of the Year 2024, for his contributions to AI research.