Privacy Protection for Life-log Video

Recent advances in wearable cameras and storage devices allow us to record the holistic human experience for an extended period of time. Such a life-log system can capture audio-visual data anywhere and at any time. It has a wide range of applications from law enforcement, journalism, medicine to personal archival. On the other hand, there is a natural apprehension towards such an intrusive system as the audio-visual information of unsuspecting subjects captured in the life-log video may be misused. Thus, along with the technical challenges, the privacy and legal issues arising in such recordings must be carefully addressed. In this paper, we describe a wearable life-log system that combines real-time audio distortion and visual blocking to protect the privacy of the subjects captured in life-log video. For audio, our system automatically isolates the subject's speech and distorts it using a pitch-shifting algorithm to conceal the identity. For video, our system uses a real-time face detection, tracking and blocking algorithm to obfuscate the faces of the subjects. Extensive experiments have been conducted on interview videos to demonstrate the ability of our system in protecting the identity of the subject while maintaining the usability of the life-log video.