Zero-Shot Learning Using Synthesised Unseen Visual Data with Diffusion Regularisation
- Submitting institution
-
University of Durham
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 124887
- Type
- D - Journal article
- DOI
-
10.1109/TPAMI.2017.2762295
- Title of journal
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Article number
- -
- First page
- 2498
- Volume
- 40
- Issue
- 10
- ISSN
- 01628828
- Open access status
- Compliant
- Month of publication
- -
- Year of publication
- 2017
- URL
-
https://doi.org/10.1109/TPAMI.2017.2762295
- 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)
-
A - Innovative Computing
- Citation count
- 25
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- When this paper was accepted, it was the first work that established a new paradigm to synthesise unseen visual data. In addition to its conference version: "From zero-shot to conventional supervised classification: unseen visual data synthesis" in CVPR 2017, more than a hundred follow-up papers have adopted this paradigm with generative models, e.g. GANs in a wide range of applications.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -