From Eyes to Face synthesis: a new approach for human-centered smart surveillance
- Submitting institution
-
University of Sunderland
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 675
- Type
- D - Journal article
- DOI
-
10.1109/ACCESS.2018.2803787
- Title of journal
- IEEE Access
- Article number
- -
- First page
- 14567
- Volume
- 6
- Issue
- 99
- ISSN
- 2169-3536
- Open access status
- Compliant
- Month of publication
- February
- Year of publication
- 2018
- URL
-
http://sure.sunderland.ac.uk/id/eprint/8956/
- 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
-
4
- Research group(s)
-
-
- Citation count
- 6
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Face recognition for personal identification is one of the most important tasks in smart surveillance e.g. for public security, criminal investigation and anti-terrorism. Currently face occlusion causes problems for automated person identification. This paper developed a new deep learning approach, based on conditional generative adversarial networks (GAN), to predict the entire face information from data for the eyes region. Empirical evaluation revealed significant benefits for the new approach, which offers a promising solution for the personal identification, even where faces are occluded.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -