ForgetMeNot: Memory-Aware Forensic Facial Sketch Matching
- Submitting institution
-
Queen Mary University of London
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2561
- Type
- E - Conference contribution
- DOI
-
10.1109/CVPR.2016.601
- Title of conference / published proceedings
- 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- First page
- 5571
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- December
- 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
-
-
- Research group(s)
-
-
- Citation count
- 19
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- We reformulate the problem of forensic sketch to image matching, to better align it with practical expectations from law enforcements. Work recognises a salient memory gap between sketches rendered by artists and the actual face seen by the eye witness. It then proposes a novel Gaussian Process Regression framework to model the time gaps. The system yields best matching performance on the largest forensic photo-sketch dataset which is also part of the NIST benchmarks. The superior performance had drawn considerable attention from the Met Police showing a strong interest to put it into real use. The work is patent pending.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -