Use Second Screen to Enhance TV Viewing Experiences

In this paper, we propose a second-screen interactive TV viewing scenario, and build an Android-based prototype to verify our concept and design. The demo app detects the TV program that a user is watching, and then dynamically pushes additional information through the second screen to the user. In this prototype, we apply audio fingerprinting technology to detect TV programs, and users do not have to manually surf hundreds of channels. In addition, face recognition and tennis event detection technologies have been employed to extract critical elements from the TV contents. With these content analysis tools, we can obtain video metadata more easily and need not rely on content providers. A usability test with 20 participants has been conducted. Our test shows participants are interested in our second screen app, and they would like more features and information from the app. Next, we plan to further refine the app’s design to enhance user experience.