Attending to Discriminative Certainty for Domain Adaptation
- Submitting institution
-
The University of Bath
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 213132907
- Type
- E - Conference contribution
- DOI
-
10.1109/CVPR.2019.00058
- Title of conference / published proceedings
- IEEE Conference on Computer Vision and Pattern Recognition
- First page
- 491
- Volume
- -
- Issue
- -
- ISSN
- 1063-6919
- Open access status
- Exception within 3 months of publication
- Month of publication
- January
- Year of publication
- 2020
- 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
- 11
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- CVPR is a top venue for computer vision research. This work demonstrated a probabilistic approach for domain adaptation, to allow deep learning models to be useful in a variety of settings for which they have not been trained. The approach suggests use of certainty based attention models could ensure that the models generalise better and we would also be able to analyse the certainty of the regions where the model is improving through the adaptation. This enhances the ability to explain the domain adaptation procedure. This is one of the first such explainable adaptation methods.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -