An Audio-Video Summarization Scheme Based on Audio and Video Analysis

AbstractThe availability of video files in the Internet is growing at an excep-tional speed and in the near future video browsing will be a commonactivity. To facilitate such activity it will be necessary to have a smallclip for any given video. Currently, video skimming and video summariza-tion techniques can reduce the temporal representation of a given video.However, most of these techniques do not include audio in the producedsummaries. Here, we propose a mechanism that, using audio and videoanalysis, produces video summaries coupled with intelligible audio. Ex-perimental results show that the summaries are largely reduced (up to50%) and that the perceived video quality may be comparable to the oneof the original video (in term of jerkiness). Consumers satisfaction hasbeen investigated through MOS and results show that our summaries canbe considered as an alternative to the original videos. 1 Introduction Thanks to the advances in networking and multimedia technologies, video infor-mation are massively entering our life. Nowadays, a variety of devices (palms,cellulars, laptops) has Internet connection (either Fiber, DSL, Cable, dial-up,UMTS, GPRS) and can easily play out video files.The popularity of video files is highlighted by the increasing number of TV-show downloads that P2P systems are experiencing and by the video searching