Deep learning based combating strategy for COVID-19 induced increased video consumption

摘要

COVID19 epidemic has brought tremendous changes globally. The adoption of prevention strategies like lockdown and remote working has suddenly changed all aspects of human life. A surge shift in online mode has been observed and it tremendously increased the internet traffic with a suitable rise in video content. Research and surveys state that these changes will become new normal and will last for a long time. As the bandwidths are limited and cannot be expanded instantly, there arose a need for alternate techniques to be explored to deal with growing video content efficiently. The lightweight and powerful deep learning based video compression and analytics techniques may help in efficiently processing video content. Deep learning based techniques are already giving potent results both in the video compression and video analytics domain independently. In this paper, the accelerated impact of COVID19 on video compression methods has been demonstrated and proposed joint video compressioncumanalytics scheme which may significantly provide fast and efficient video analytics from the compressed video optimizing whole network.