Nested Shallow CNN-Cascade for Face Detection in the Wild
- Submitting institution
-
Swansea University / Prifysgol Abertawe
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 32108
- Type
- E - Conference contribution
- DOI
-
10.1109/FG.2017.29
- Title of conference / published proceedings
- 2017 IEEE 12th International Conference on Automatic Face & Gesture Recognition
- First page
- 165
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- June
- Year of publication
- 2017
- URL
-
http://csvision.swan.ac.uk/uploads/Site/Publication/jd17fg.pdf
- 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
- 2
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The use of machine learning in face recognition in the wild is challenging, technically and socially. Instead of using increasingly complex and deep neural network structures, this work presents a novel approach based on nested CNN-cascade learning with shallow network architectures that allow efficient and progressive elimination of negative hypothesis from easy to hard via self-learning discriminative representations from coarse to fine scales. This approach differs significantly from current trends and offers an effective alternative that is is particularly appealing to real time applications. Experiments on several datasets with various real-world conditions show state-of-the-art performance of the proposed method.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -