The Community for Technology Leaders
ABSTRACT
We introduce a new large-scale video dataset designed to assess the performance of diverse visual event recognition algorithms with a focus on continuous visual event recognition (CVER) in outdoor areas with wide coverage. Previous datasets for action recognition are unrealistic for real-world surveillance because they consist of short clips showing one action by one individual [15, 8]. Datasets have been developed for movies [11] and sports [12], but, these actions and scene conditions do not apply effectively to surveillance videos. Our dataset consists of many outdoor scenes with actions occurring naturally by non-actors in continuously captured videos of the real world. The dataset includes large numbers of instances for 23 event types distributed throughout 29 hours of video. This data is accompanied by detailed annotations which include both moving object tracks and event examples, which will provide solid basis for large-scale evaluation. Additionally, we propose different types of evaluation modes for visual recognition tasks and evaluation metrics along with our preliminary experimental results. We believe that this dataset will stimulate diverse aspects of computer vision research and help us to advance the CVER tasks in the years ahead.
INDEX TERMS
surveillance video, large-scale video dataset, diverse visual event recognition algorithm, continuous visual event recognition, outdoor scenes, moving object tracks, evaluation metrics, computer vision, CVER tasks
CITATION

Jong Taek Lee et al., "A large-scale benchmark dataset for event recognition in surveillance video," CVPR 2011(CVPR), Providence, RI, 2011, pp. 3153-3160.
doi:10.1109/CVPR.2011.5995586
531 ms
(Ver 3.3 (11022016))