Training an adaptive dialogue policy for interactive learning of visually grounded word meanings
- Submitting institution
-
Heriot-Watt University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 24184147
- Type
- E - Conference contribution
- DOI
-
10.18653/v1/W16-3643
- Title of conference / published proceedings
- Proceedings of the SIGDIAL 2016 Conference
- First page
- 339
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- September
- Year of publication
- 2016
- 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
- The first work combining a formal Natural Language grammar with Computer Vision & Machine Learning to enable natural, multi-modal conversational agents that learn new concepts from interaction. The work has had some international impact as evidence by citations by other groups. It currently forms the basis of an emerging proposal as well as a journal article. The model has the potential to have very significant impact. Top Conference in the Dialogue Systems (Conversational Agents) area (SIGDIAL).
Contribution: Initial ideas, Model, Implementation, Evaluation, Writing.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
