Conference proceeding
Soft constrained leading voice separation with music score guidance
The 9th International Symposium on Chinese Spoken Language Processing, pp.565-569
09/2014
DOI: 10.1109/ISCSLP.2014.6936723
Abstract
Separating leading voice from a music mixture remains challenging for automatic systems. Competing harmonics from music accompaniment severely interfere the leading voice estimation. To properly extract the leading voice, separation algorithms based on source-filter modeling of human voice and non-negative matrix factorization have been introduced. This paper extends this approach with a statistical weighting scheme to rank various pitch candidates with music score information. It imposes a soft constraint on the likelihood of these pitch candidates, so the interference from music accompaniment on leading voice estimation is reduced. Our experiments showed that this soft-constrained separation with score guidance provides accurate inference about the leading vocal pitch with reliable score and remains robust for erroneous score.
Details
- Title: Subtitle
- Soft constrained leading voice separation with music score guidance
- Creators
- Renbo Zhao - National University of SingaporeS. W. Lee - Institute for Infocomm ResearchDong-Yan Huang - Human Language Technol. Dept., ASTAR, Singapore, SingaporeMinghui Dong - Institute for Infocomm Research
- Resource Type
- Conference proceeding
- Publication Details
- The 9th International Symposium on Chinese Spoken Language Processing, pp.565-569
- Publisher
- IEEE
- DOI
- 10.1109/ISCSLP.2014.6936723
- Language
- English
- Date published
- 09/2014
- Academic Unit
- Business Analytics
- Record Identifier
- 9984446554602771
Metrics
1 Record Views