Robust multi-speaker tracking via dictionary learning and identity modeling
- Submitting institution
-
Kingston University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11-15-1349
- Type
- D - Journal article
- DOI
-
10.1109/TMM.2014.2301977
- Title of journal
- IEEE Transactions on Multimedia
- Article number
- -
- First page
- 864
- Volume
- 16
- Issue
- -
- ISSN
- 1520-9210
- Open access status
- Out of scope for open access requirements
- Month of publication
- -
- Year of publication
- 2014
- 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
-
-
- Research group(s)
-
-
- Citation count
- 22
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Identifying and tracking speakers in smart environments is an extremely challenging task. This paper combines several learning and modelling methods to address it. It is significant in being one of the first uses of existing machine methods in the tracking process. This concept of combined learning and tracking has become one of the common approaches used in deep learning-based tracking methods. Moreover, the idea of using speaker identity in the tracking process has since been adopted by many deep learning tracking applications. Additionally, the COVID-19 pandemic which has increased needs for remote working and conferencing, increasing this work's significance.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -