Research project

COHABIT

Clinical assessment of behavior often relies on self-reporting, interviews, and observations. Those methods pose challenges for reliability, validity, and scalability. AI can help with that, but it requires large amounts of correctly labelled data for training.

research-project-marsai

COHABIT

This project will develop the knowledge and tools needed to solve this problem. The PhD student will carry out research to develop a combination of active learning and interface design to enable rapid upscaling of computer vision-based tracking systems in clinical domains. We aim to provide a flexible and powerful tool for smart annotation of human behavior that will become an indispensable component for clinical applications requiring human behavioral analysis.

The consortium

The project is led by Noldus Information Technology and the other partner is the Department of Information and Computing Sciences of the University of Utrecht.

The role of Noldus

As the project lead, Noldus will employ the PhD student and so is responsible for the main deliverables of the project.

Funding

This collaboration project [HH-PPS-25020-PhD] is co-funded by PPP Subsidy awarded by Health~Holland, Top Sector Life Sciences & Health, in collaboration with Noldus and the University of Utrecht to stimulate public-private partnerships.

ALT 2