• DocumentCode
    892601
  • Title

    Detecting Ads in Video Streams Using Acoustic and Visual Cues

  • Author

    Covell, Michele ; Baluja, Shumeet ; Fink, Michael

  • Volume
    39
  • Issue
    12
  • fYear
    2006
  • Firstpage
    135
  • Lastpage
    137
  • Abstract
    A new technique provides accurate, efficient ad detection for online TV rebroadcasts. As online television grows in popularity, providers are seeking cost-effective ways to replace original advertisements in rebroadcasts with new ads that are not only fresher but that can target individual viewers´ interests and preferences. Alternatively, completely removing commercials might be desirable for subscribers who pay for reused content. Researchers have developed various heuristic techniques to detect and remove ads in video streams. Many of these exploit common differences between advertising and program material including cut rates, soundtrack volume, and surrounding black frames
  • Keywords
    acoustic signal processing; advertising; digital video broadcasting; video signal processing; video streaming; acoustic cues; advertisement detection; online TV rebroadcast; video stream; visual cues; Acoustic signal detection; Audio databases; Filters; Hamming distance; Monitoring; Music information retrieval; Sampling methods; Signal processing; Streaming media; Visual databases; Entertainment computing; Invisible computing;
  • fLanguage
    English
  • Journal_Title
    Computer
  • Publisher
    ieee
  • ISSN
    0018-9162
  • Type

    jour

  • DOI
    10.1109/MC.2006.421
  • Filename
    4039265