Output details
15 - General Engineering
University of Hertfordshire
Article title
H-Semantics : A Hybrid Approach to Singing Voice Separation
Type
D - Journal article
DOI
-
Title of journal
Journal of the Audio Engineering Society
Article number
-
Volume number
60
Issue number
10
First page of article
831
ISSN of journal
1549-4950
Year of publication
2012
Number of additional authors
3
Additional information
- This paper details the extensive research conducted into the extraction of a singing voice from stereophonic studio recordings.
- Building on techniques developed previously at the University of Hertfordshire, this novel approach introduced a major step forward in automated singing voice separation, with such important applications as lyrics recognition, singer identification, and melody extraction.
- The proposed method, uses the integration of time- and frequency-domain processes for the segregation of singing voice from its music accompaniment, and is shown to offer superior performance over the previously reported techniques.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
