Video Concept Detection

Principal Investigator(s): 
Gerald Friedland

Massive numbers of video clips are generated daily on many types of consumer electronics and uploaded to the Internet. In contrast to videos that are produced for broadcast or from planned surveillance, the "unconstrained" video clips produced by anyone who has a digital camera present a significant challenge for manual as well as automated analysis. Such clips can include any possible scene and events, and generally have limited quality control. In ALADDIN, we aim to develop methods for finding videos that contain specified concepts based on the analysis of audio and video content out of a large dataset of consumer-produced videos. Example concepts are "One or more people make a cake," "Batting a run in," and "Assembling a shelter." The project is a collaboration between the Audio and Multimedia Group, SRI (Sarnoff), Carnegie-Mellon University, and the University of Central Florida.

Funding is provided by IARPA and Lawrence Livermore National Laboratory.