Title:
音声信号の基礎となるメロディの抽出
Document Type and Number:
Japanese Patent JP2008516288
Kind Code:
A
Abstract:
According to the invention, melody extraction or automatic transcription can be made significantly more stable and possibly also less expensive by sufficiently taking into account the assumption that the main melody is the portion of a piece of music which is perceived by humans as being the loudest and most prominent. Hence, in order to determine the melody of the audio signal, a melody line which extends across the time/spectral representation is determined (780) first by unambiguously assigning exactly one spectral component or a frequency bin of the time/spectral representation to each time slot or frame, i.e. according to a special embodiment of the invention, the spectral component that results in the sound having the maximum intensity on said frame.
Inventors:
Straighten bearer frank
Weiss Martin
Derboven class
Kramer Marx
Weiss Martin
Derboven class
Kramer Marx
Application Number:
JP2007536025A
Publication Date:
May 15, 2008
Filing Date:
September 23, 2005
Export Citation:
Assignee:
Fraunhofer-Gesellschaftsza Felderung der Angewanten Vorschung A. Fao
International Classes:
G10G3/04; G10H1/00; G10L25/18; G10L25/51
Domestic Patent References:
| JPH1195780A | 1999-04-09 | |||
| JP2003173195A | 2003-06-20 | |||
| JPH06291671A | 1994-10-18 | |||
| JPH04220880A | 1992-08-11 | |||
| JP2000099009A | 2000-04-07 | |||
| JP2003022096A | 2003-01-24 | |||
| JPH1195753A | 1999-04-09 | |||
| JP2002023799A | 2002-01-25 |
Attorney, Agent or Firm:
Masahiro Okada