Fine-Grained Action Retrieval Through Multiple Parts-of-Speech Embeddings
- Submitting institution
-
University of Bristol
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 219667347
- Type
- E - Conference contribution
- DOI
-
10.1109/ICCV.2019.00054
- Title of conference / published proceedings
- 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
- First page
- 450
- Volume
- -
- Issue
- -
- ISSN
- 2380-7504
- Open access status
- Compliant
- Month of publication
- November
- Year of publication
- 2019
- 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
-
3
- Research group(s)
-
C - Visual Information Lab
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper exploits Part-of-Speech (PoS) in captions as supervision for fine-grained video retrieval, jointly learning multiple embedding spaces. The work resulted in the award of a PhD to first author Wray. Part of a collaboration with Naver Labs Europe via Damen’s EPSRC First Grant LOCATE (EP/N033779/1). This collaboration continues with PhD student (Munro) ongoing an internship with co-authors Larlus and Csurka. Damen has presented this work in multiple keynotes [WILV@CVPR2020, CAMP@ECCV2020, HUMA@MM2020]. A new challenge on Multi-Instance Retrieval emerged from this work [https://competitions.codalab.org/competitions/26138]. Follow-up work by Damen [Dought et al, CVPR 2020 Action Modifiers] extended this to “adverbs” PoS
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -