
Research Article
Synopsis of Video Files Using Neural Networks: Component Analysis
@INPROCEEDINGS{10.1007/978-3-031-17292-2_2, author={Georgi Kostadinov}, title={Synopsis of Video Files Using Neural Networks: Component Analysis}, proceedings={Computer Science and Education in Computer Science. 18th EAI International Conference, CSECS 2022, On-Site and Virtual Event, June 24-27, 2022, Proceedings}, proceedings_a={CSECS}, year={2022}, month={11}, keywords={Video synopsis Convolutional neural networks Machine learning Object localization Multiple object tracking Feature extraction Background segmentation Person re-identification}, doi={10.1007/978-3-031-17292-2_2} }
- Georgi Kostadinov
Year: 2022
Synopsis of Video Files Using Neural Networks: Component Analysis
CSECS
Springer
DOI: 10.1007/978-3-031-17292-2_2
Abstract
The following paper provides a detailed analysis of the components of a novel framework for generating synopsis of CCTV videos. A synopsis is a video file obtained by overlaying the main objects from a source video on a single scene. This allows for a file length reduction and optimization of the storage of such files. This paper extends the presented work based on convolutional neural networks by discussing the effect that such algorithms may have on the final synopsis result which in turn helps in understanding how they can be further improved for this task. For the purposes of the component analysis presented in this paper, specialized datasets and metrics were selected to quantify the quality of the algorithms of the video synopsis framework.