Learning graphs to model visual objects across different depictive styles
- Submitting institution
-
The University of Bath
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 146390876
- Type
- E - Conference contribution
- DOI
-
10.1007/978-3-319-10584-0_21
- Title of conference / published proceedings
- Computer Vision – ECCV 2014 : 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VII
- First page
- 313
- Volume
- 8695
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- September
- 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
-
2
- Research group(s)
-
-
- Citation count
- 4
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Recognising objects across depictions (photos, paintings, etc) is a major open problem with applications in internet search, database tagging, cultural history, assistive computing and elsewhere. So far as we know, only this approach generalises stabley across depictions. The database included in the paper has been used by researchers across the globe including world leading groups (e.g. University of Washington, Universite Paris-Saclay, University of Pittsburgh). The technique is central to our relationship with Nankai University, with whom we have a CSC student. It underpins current work with Art Historians, Disney, Ninja Theory, RNIB and conversations with the British Museum.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -