Case Study: Teton.ai prevents falls in care homes and hospitals with Encord

Using Computer Vision to Prevent Falls in Care Homes and Hospitals

Teton.ai, a company creating computer vision tools for fall prevention in healthcare, faced significant challenges in training its AI models. They needed a flexible and efficient way to annotate large volumes of video data to teach their models to identify pre-fall behaviors in patients, moving beyond the limitations of simple bounding box annotations. To achieve this, they turned to Encord and its training data platform.

Encord provided the solution with its adaptable annotation platform and SDK. This allowed Teton.ai to implement a more effective labeling strategy, directly classifying patient states and behaviors. The Encord SDK enabled a streamlined workflow for pre-labeling, manual review, and model training. This partnership gave Teton.ai the flexibility to pivot their approach, resulting in significantly improved model performance for their fall-prevention system, which is now deployed in hospitals and care homes.

Teton.ai