Unsupervised learning methods for source separation in monaural music signals

Signal Processing Methods for Music Transcription

Volumn , Issue , 2006, Pages 267-296

  Virtanen, Tuomas ^a  

^a TAMPERE UNIVERSITY OF TECHNOLOGY   (Finland)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 64849096680     PISSN: None     EISSN: None     Source Type: Book    
DOI: 10.1007/0-387-32845-9_9     Document Type: Chapter

References (4)

1. The concept of probability distribution function is described in Chapter 2.
2. Singular value decomposition can also be used to estimate the number of components [73].
3. ICA aims at maximizing the independence of the output variables, but it cannot guarantee their complete independence, as this depends also on the input signal.
4. This approach is related to the fundamental frequency estimation method of Brown, who calculated the cross-correlation between an input spectrum and a single harmonic template on the logarithmic frequency scale [54].