An Automatic Event Recognition System for Real-Time Streaming Systems Using Transfer Learning Based on Convolutional Neural Networks
Jih‐Sheng Chang, Hsiu‐Mei Chou, Wei-Chih Su, Fang‐Pang Lin, Yi-Hao Hsiao · 2018
The demand for live streaming is on the rise and keeps on growing every single day. A wide range of streaming applications such as live broadcast apps, online TV shows, and live surveillance systems, are thriving more than ever before. To ensure the stable delivery of streaming without suffering from interruptions on account of camera failure is significant for an unmanned surveillance system. In this paper, we are concerned with the development of an automatic event recognition model for a real-time streaming system using transfer learning based on Convolutional Neural networks, which is able to detect various types of abnormal events from live streaming so as to improve the overall system reliability. The performance has been verified through applying experiments on a real-time streaming system to demonstrate the practicability of the proposed model.