A Time-frequency Masking Based Random Finite Set Particle Filtering Method for Multiple Acoustic Source Detection and Tracking
- Submitting institution
-
University of Edinburgh
(joint submission with Heriot-Watt University)
- Unit of assessment
- 12 - Engineering
- Output identifier
- 41594783
- Type
- D - Journal article
- DOI
-
10.1109/TASLP.2015.2479041
- Title of journal
- IEEE Transactions on Audio, Speech and Language Processing
- Article number
- -
- First page
- 2356
- Volume
- 23
- Issue
- 12
- ISSN
- 1558-7916
- Open access status
- Out of scope for open access requirements
- Month of publication
- September
- Year of publication
- 2015
- 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)
-
C - SSS
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper provides an algorithm for detecting and tracking multiple concurrent acoustic talkers using a microphone array. It introduces a novel method to improve the time-frequency masking approach. Such algorithms have application in distant speech recognition for voice assistance, human tracking systems, and other target detection modalities. This work has been adopted by the acoustic tracking community (e.g. doi:10.1109/TASLP.2016.2590146, doi:10.1109/IranianCEE.2016.7585675). This work is an extension of a major output from Zhong’s PhD and non-concurrent tracking work (doi:10.1016/j.sigpro.2013.09.002). This work forms the basis for localisation techniques in DSTL/UDRC EP/K014277/1 (£4.3M).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
