Event-Camera Technology for privacy-secured sensing

four frames showing event camera output

A workshop was organized in November 2022 to introduce the new and exciting event camera technology for privacy-secured sensing, a topic aligned with the themes of the ADAPT Special Interest Group (SIG) on "Attentive AI". Event cameras capture pixel-level changes in image intensity that can be used to sense changes in an imaged scene without creating a human-understandable image or sequence of video frames. This enables sensing without creating a permanent image/video record.

Thus, when paired with neural-network processing, event cameras offer an excellent emerging technology for human-computer interaction, while protecting users from data privacy concerns. Furthermore, the benefits of event cameras extend to computational efficiency, as demonstrated in this figure. 

image showing a demo of event camera's computational efficiency

Several technical presentations covering some novel use cases of the event camera technology in the context of automotive applications were made at the workshop, as well as a live demo of a Time Event Driver Monitoring System. Links to video recordings and/or presentations from all workshop sessions/presentations are provided here:

Research Presentations

Workshop date:  15th November 2022
Host: Xperi, Galway
Organized by: Prof. Peter Corcoran of University of Galway (www.universityofgalway.ie/c3i/) and Dr. Joseph Lemley of Xperi - Xperi Driver Monitoring Systems (DMS) (www.xperi.com)