Wednesday, April 22, 15:00 - 17:30, Location: Show and Tell Area B
Eric Zavesky and Shih-Fu Chang
Columbia University's CuZero video search engine is a novel system that combines a guided stage of query formulation and a real-time interaction stage with intuitively organized video content. It demonstrates unique capabilities in mapping user's search targets into meaningful queries, and simultaneously manipulating queries and results in a zero-latency fashion. The current prototype supports a diverse pool of video search tools, including concept-based search over more than 400 concept models and query by example tools using different visual features. The system has been tested in TRECVID 2008 and Video Olympics evaluations over test sets of more than 150 hours, 120,000 shots of digital video content. Live demos will be shown on site.