| 摘要 |
<p>An audio and video synchronizing perceptual model 104 associates relative emotional impact with different audio portions (eg. music with or without speech) in order to determine transition points 204 (of eg. loud, rhythmic music to soft, slow music) facilitating automatic synchronization 206 of audio data to video data. Perceptual characteristics (eg. a filtered, normalized spectrogram which indicates a change in harmonic content over time) may be compared from different portions in order to identify transitions and sync them to video transitions.</p> |
