Music Signal Processing Exploiting Spectral Fluctuation of Singing Voice Using Harmonic/Percussive Sound Separation
H. Tachibana · Institutional Repositories DataBase (IRDB) · 2014
Singing voice is one of the most impressive components in music signals.Extracting singing voice from mixed music signals has significance because of the potential needs for contentbased music search and potential applicability to a technical component of interactive music players.The thesis contains the descriptions of novel music signal processing techniques on singing voice, especially singing voice enhancement, pitch estimation of singing melody, automatic real-time audio-to-audio karaoke generation system (singing voice suppression), 1 UT (-03/2013)