Binarized convolutional landmark localizers for human pose estimation and face alignment with limited resources
- Submitting institution
-
University of Nottingham, The
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1330525
- Type
- E - Conference contribution
- DOI
-
10.1109/ICCV.2017.400
- Title of conference / published proceedings
- 2017 IEEE International Conference on Computer Vision (ICCV 2017)
- First page
- 3726
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- October
- Year of publication
- 2017
- 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
-
1
- Research group(s)
-
-
- Citation count
- 29
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Deployment of deep-learnt models is an increasingly difficult challenge, as models continue to grow in size and complexity. This approach reduces the model complexity while retaining a similar level of performance, enabling deployment to slower devices such as mobile phones, where compute and power is limited. For human pose and face alignment, this may enable real-time and interactive use on such devices for a wide range of applications. The work has led to numerous follow up publications, in particular by commercial organisations. In particular, Samsung AI published at least five papers building on this work.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
