AgeDB: the first manually collected, in-the-wild age database
- Submitting institution
-
Middlesex University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 794
- Type
- E - Conference contribution
- DOI
-
10.1109/CVPRW.2017.250
- Title of conference / published proceedings
- 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
- First page
- 1997
- Volume
- -
- Issue
- -
- ISSN
- 2160-7516
- Open access status
- Compliant
- Month of publication
- July
- Year of publication
- 2017
- URL
-
http://eprints.mdx.ac.uk/22044/
- 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
-
5
- Research group(s)
-
-
- Citation count
- 30
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- One of the most challenging and important tasks in Computer Vision is face recognition with the change of ages. While there exist datasets annotated with age attributes, these datasets are semi-automatically collected and thus contain noisy labels. This paper is significant because it details the first database, AgeDB, that contains images manually annotated with accurate to the year, noise-free labels. The paper describes experiments using state-of-the-art algorithms that are trained using AgeDB. The data makes the algorithms suitable forage-invariant face verification, age estimation and face age progression “in-the-wild”.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -