Human Perception-Enhanced Camera System for Web Conferences Leveraging Device Motions

Anish Shrestha,Zeyu Deng,Chen Wang
DOI: https://doi.org/10.1145/3447993.3510591
2021-01-01
Abstract:We present a demonstration of a human perception-enhanced camera system for web conferencing that protects the user's privacy. Given that people easily forget about their active camera during web conferences, the system advertises the camera's active status via its motions to remind users that they are being watched by others. This prevents inadvertent privacy leakage. The system is developed based on a motorized camera, which moves according to the user's head coordinates just like an eye is looking at the user's face in front of the desk rather than remotely or virtually. The basic idea is to exploit the original human body sense of environmental motions for human-camera interaction, which does not require looking straight at the camera or its LED light to actively check its status. In this demonstration, we showcase our implementation of the human perception-enhanced camera system and invite participants to use the system for web conferences (e.g., Zoom and Google Hangout), which illustrates the system's ability to extend the virtual social interaction to the physical world and the effectiveness of using the camera motion as a non-intrusive awareness indicator.
What problem does this paper attempt to address?