AI for Health and Human Performance
Current clinical assessments often rely on short observations, subjective scores, or specialized equipment. We develop interpretable, video-based biomarkers that quantify how an individual moves, changes over time, and responds to intervention.
This impact area applies our human-modeling and biomechanics research to health, mobility, and human performance. We build interpretable systems — grounded in biomechanics and explained in language clinicians and patients can act on — for remote, unobtrusive, and longitudinal assessment.
How this differs from Physical Human Intelligence: Physical Human Intelligence asks how machines can understand human bodies and movement. AI for Health and Human Performance applies those capabilities to mobility, rehabilitation, disease monitoring, coaching, and personalized assessment.
What we work on
- Video-based digital biomarkers
- Clinical gait and mobility assessment
- Longitudinal tracking of disease and recovery
- Personalized movement analysis
- Rehabilitation monitoring
- Biomechanics-grounded clinical AI
- Physical coaching and skill assessment
- Human performance and readiness
- Remote and unobtrusive health assessment
- Interpretable vision-language systems for clinicians and patients