Public Security Video and Image Analysis Challenge: A Retrospective
The Public Security Video and Image Analysis Challenge (PSVIAC) is a benchmark in object detection and instance search on public security surveillance videos. This challenge is first held in 2016, attracting participation from more than twenty institutions. This paper provides a review of this challenge, including tasks definition, datasets creation, ground truth annotation, and results comparison and analysis. We conclude the paper with some future improvements.
KeywordsGround Truth Object Detection Object Class Vehicle Detection Occlusion Condition
Our research was sponsored by following projects: Program of Science and Technology Commission of Shanghai Municipality (No. 15530701300, 15XD1520200, 14DZ2252900); 2012 IoT Program of Ministry of Industry and Information Technology of China; Key Project of the Ministry of Public Security (No. 2014JSYJA007); Shanghai Science and Technology Innovation Action Plan (No. 16511101700).
- 3.Over, P., Fiscus, J., Sanders, G., et al.: Trecvid 2014-an overview of the goals, tasks, data, evaluation mechanisms and metrics. In: Proceedings of TRECVID (2014)Google Scholar
- 4.Patino, L., Ferryman, J.: Pets: dataset and challenge. In: IEEE International Conference on Advanced Video and Signal Based Surveillance 2014, pp. 355–360 (2014)Google Scholar
- 6.Zhu, M.: Recall, precision, and average precision. Department of Statistics and Actuarial Science, University of Waterloo (2004)Google Scholar
- 7.Turpin, A., Scholer, F.: User performance versus precision measures for simple search tasks. In: Proceedings of the International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 11–18 (2006)Google Scholar