Protoype of AI-based Smart-Space

Description

As part of the expertise in human-centred data acquisition of the ISIA LAB (Numediart Institute), the goal of the post-doc fellowship will be to produce an adaptative smart sensing space technological and research brick. This brick will be able to provide:

1) The detection and tracking of objects and people position and pose based on a plurality of RGB-D cameras (such as the ZED2 camera from Stereolabs or the OAK-D and OAK-D light cameras from OpenCV). This data is calibrated and synchronized into a single common space

2) From people data extract high-level information such as actions recognitions, social signal and “abnormal” or “unpredictable” situations

In addition to the research challenge, a technological challenge will be to provide a adaptative framework which is able to manage algorithms running on a common server but also algorithms which might run on lighter computation setups on the edge especially on cameras.

This brick can be adapted to the numerous technological projects that we have with different companies and universities but also to human-sciences projects that we have with psychology partners.