Acoustic adaptation to dynamic background conditions with asynchronous transformations
- Submitting institution
-
The University of Sheffield
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2436
- Type
- D - Journal article
- DOI
-
10.1016/j.csl.2016.06.008
- Title of journal
- Computer Speech and Language
- Article number
- -
- First page
- 180
- Volume
- 41
- Issue
- -
- ISSN
- 0885-2308
- Open access status
- Compliant
- Month of publication
- July
- Year of publication
- 2016
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
1
- Research group(s)
-
G - Speech and Hearing
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Speech in real world conditions is often mixed with other background signals. This paper introduced a novel method for unsupervised detection of asynchronous background conditions during the recognition process, delivering one of the core objectives of the EPSRC programme grant EP/I031022/1. Follow-up work demonstrated its effect in the MGB competitions (http://www.mgb-challenge.org/): improved media speech recognition, automatic detection of media genres, and improved subtitle alignment (see https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=7404863). The asynchronous latent factor approach was a precursor to industry sponsored research (Huawei Innovation Research Program (HIRP): X/159898-11) on additive demixing of signals properties.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -