Easily deployable multi-camera system for indoor building occupant behaviour monitoring

Supervisors: Julien Nembrini

Student: TBD

Project status: Open

Year: 2024

Using pose-estimation models such as OpenPifPaf, the project will merge multiple cameras to cope with typical indoor occultation (possibly based on calliscope). Focus will be put on the ease of deployment, data privacy and the communication of sensed data to participants. Further work may include activity categorization based on pose information.

This project can be a taken as Master or Bachelor project.