Mukesh Saini

Privacy-Aware Surveillance System Design

Video surveillance is a very effective means of monitoring activities over a large area with cameras as extended eyes. However, this additional security comes at the cost of privacy loss of the citizens not involved in any illicit activities. Because an adversary can use prior knowledge to infer the identities of individuals in the video even in the absence of facial information, we develop a privacy-aware surveillance framework in which we identify the implicit channels of identity leakage, quantify the privacy loss through non-facial information, and propose solution to block these channels for near zero privacy loss with minimal utility loss. The proposed privacy loss model considers facial as well as non-facial information and is able to consolidate the identity leakage through multiple events and multiple cameras. Privacy loss is modelled as an adversary’s ability to correlate sensitive information to the identity of the individuals in the video. Anonymity based approach is used to consolidate the identity leakage through explicit channels of bodily cues such as facial information; and other implicit channels that exist due to ‘what’, ‘when’ and ‘where’ information.

Moreover, any privacy preserving method usually affects the utility of the data; therefore, the choice of data transformation is paramount to ensure an acceptable tradeoff between the privacy and the utility. We propose utility models and privacy preservation framework for the applications of video surveillance and video data publication. Through experiments it is found that current privacy protection methods include high risk of privacy loss while the proposed framework provides more robust privacy loss measures and better tradeoff of security and privacy.

For more details, please email: sainimukesh@gmail.com