A Bag of Expression framework for improved human action recognition
- Submitting institution
-
Kingston University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11-43-1328
- Type
- D - Journal article
- DOI
-
10.1016/j.patrec.2017.12.024
- Title of journal
- Pattern Recognition Letters
- Article number
- -
- First page
- 39
- Volume
- 103
- Issue
- -
- ISSN
- 0167-8655
- Open access status
- Compliant
- Month of publication
- -
- 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
-
-
- Research group(s)
-
-
- Citation count
- 14
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Bag of Expression (BoE) is an extension of the Bag of Words (BoW) method for Human Action Recognition from video data, where spatio-temporal contextual relationships between words are modelled. Such enhancement offers greater view independence and more tolerance to occlusion. Experiments conducted on a variety of public benchmark datasets showed that the BoE approach delivers state-of-the-art performance, outperforming both the standard BoW and deep learning methods. This research is particularly significant because, not only is BoW a standard computer vision approach, but also it can be combined with deep learning algorithms, e.g., Bag of Deep Features.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -