Automatically Detecting Bystanders in Photos to Reduce Privacy Risks

Learn about an innovative approach to protect bystander privacy in photographs taken in public spaces. Explore a 16-minute IEEE conference talk that delves into the challenges posed by unintentional capture of individuals in photos and the potential privacy risks associated with online sharing and advanced computer vision technologies. Discover how researchers are developing methods to automatically identify bystanders in images using visual information, without requiring proactive measures from individuals. Examine the study's methodology, including an online user survey to understand human classification of subjects and bystanders, and the development of classifier models based on intuitive concepts. Gain insights into the model's performance, achieving high detection accuracy across different datasets, and its potential implications for enhancing privacy protection in the digital age.