Nonnegative features of spectro-temporal sounds for classification
SCIE
SCOPUS
- Title
- Nonnegative features of spectro-temporal sounds for classification
- Authors
- Cho, YC; Choi, SJ
- Date Issued
- 2005-07-01
- Publisher
- ELSEVIER SCIENCE BV
- Abstract
- A parts-based representation is a way of understanding object recognition in the brain. The nonnegative matrix factorization (NMF) is an algorithm which is able to learn a parts-based representation by allowing only non-subtractive combinations [Lee, D.D., Seung, H.S., 1999. Learning the parts of objects by non-negative matrix factorization. Nature 401, 788-791]. In this paper we incorporate a parts-based representation of spectro-temporal sounds into the acoustic feature extraction, which leads to nonnegative features. We present a method of inferring encoding variables in the framework of NMF and show that the method produces robust acoustic features in the presence of noise in the task of general sound classification.. Experimental results confirm that the proposed feature extraction method improves the classification performance, especially in the presence of noise, compared to independent component analysis (ICA) which produces holistic features. (c) 2004 Elsevier B.V. All rights reserved.
- Keywords
- acoustic feature extraction; general sound recognition; nonnegative matrix factorization
- URI
- https://oasis.postech.ac.kr/handle/2014.oak/24570
- DOI
- 10.1016/j.patrec.2004.11.026
- ISSN
- 0167-8655
- Article Type
- Article
- Citation
- PATTERN RECOGNITION LETTERS, vol. 26, no. 9, page. 1327 - 1336, 2005-07-01
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.