Learning to count objects in natural images for visual question answering
- Submitting institution
-
University of Southampton
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 49846430
- Type
- E - Conference contribution
- DOI
-
-
- Title of conference / published proceedings
- International Conference on Learning Representations
- First page
- 1
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- February
- Year of publication
- 2018
- 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
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper introduced a novel architecture to count objects for the very well known Visual Question Answering challenge (https://tinyurl.com/wpx6jl7). This architecture has been used by almost all subsequent state-of-the-art deep learning machines tackling this problem (e.g https://tinyurl.com/rutbc4t, this module is known as “counter” in this paper). The paper illustrates that it is possible to architecture high level cognitive units in deep leaning networks by a rigorous exploitation of prior constraints.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -