Automated Content Based Video Retrieval.
Srđan Zagorac, Ainhoa Llorente, Suzanne Little, Haiming Liu, Stefan Rüger · 2009
Abstract. We describe our experiments for the search task. Eight runs were submitted, all of them corresponding to the fully automated mode, without human interaction in the loop. The system was based on determining the distance from a query image to a pre-indexed collection of images to build a list of results ordered by similarity. We used four different metric measures and two different data normalisation approaches in our runs. We found that the results for all of the runs roughly match the median results achieved in this year’s competition. 1 Search Task The main goal of the search task (SE) is to model a person searching for video segments that contain persons, objects, events, locations, etc. of interest. These elements may be peripheral or accidental to the original subject of the video. The task consists of 24 topics (multimedia statement of information need), and the common shot boundary reference for a search test collection. A ranked list of at most 1,000 common reference shots that best satisfy the need are to be returned from the test collection. The data set used was provided by The Netherlands Institute for Sound and Vision and is a collection of MPEG-1 videos divided into 100 hours of videos for development and 280 hours for test purposes. The collection of videos is divided into shots according to the provided master shot boundary reference and selected