Audio Information Retrieval

• Blum, et al. " Audio Analysis for Content-Based Retrieval" 1996
• Acoustic features are extracted from audio database by use of digital signal processing
• Extracted features are : pitch, loudness, brightness, bandwidth, harmonicity,
• Query models and examples:
  • query by value: let p1 and p2 be two features, "retrieve all the sounds with p1 greater than 0.9 and p2 smaller than 0.2" or Select * where p1>0.9 and p2<0.2
  • query by example: "Retrieve all the sounds similar to this sound with respect to feature p1" or Select * where | this.p1 - p1 | < delta_p1
  • organization / browsing: "Sort the current sounds by feature p2" "Group the current sounds by feature p3 and p5"

前一張

下一張

回到第一張

瀏覽圖案部份