Balancing Awareness and Privacy
an always-on video link by manipulating image quality

In this project, we continue our research on how computers can support social awareness and light-weight communication between collaborators while mitigating privacy concerns. We take the approach of providing an always-on video system, where we dynamically adjusting the quality of the video and audio stream as a function of people's mutual proximity to the video camera and monitor [see related project]. As in real life, we envision a system where people can see and hear each other in high fidelity when they are close together. As they move apart, the fidelity of the image and audio decreases smoothly: while people can still garner an sense of another's availability, they cannot see details that would otherwise compromise privacy.

We will implement and test several methods for adjusting the quality of service of a video and audio stream. These systems will be controlled by proximity sensors, so that people's implicit actions as they move through their environment will dictate what they can see and hear on this audio-video channel. We will analyze how these technique, or combination of techniques, afford awareness information while mitigating privacy.

The images show a few ways that we are manipulating the running video. Another manipulation effect (not shown) is the frame rate, where frame rate increases with proximity. Audio manipulation effects will include gain control and masking transformations.

Primary Investigators

Michael Boyle (Honors Thesis)
Saul Greenberg (Supervisor)

Milestones

• Video Control System and API has been built, where we can control blur, contrast, resolution and frame rate of a live video feed from our groupware toolkit. The system runs within Windows.

Current Status

• The work is underway as part of an honors thesis project by Michael Boyle.