Fig. 12From: A distributed Content-Based Video Retrieval system for large datasetsThe detailed accuracy by class for the Hollywood2 datasetBack to article page