3D Motion Reconstruction from 2D Motion Data Using Multimodal Conditional Deep Belief Network
- Submitting institution
-
Staffordshire University
- Unit of assessment
- 12 - Engineering
- Output identifier
- 6787
- Type
- D - Journal article
- DOI
-
10.1109/ACCESS.2019.2904117
- Title of journal
- IEEE ACCESS
- Article number
- -
- First page
- 56389
- Volume
- 7
- Issue
- -
- ISSN
- 2169-3536
- Open access status
- Compliant
- Month of publication
- May
- Year of publication
- 2019
- URL
-
http://ieeexplore.ieee.org/document/8707086
- 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)
-
B - Centre for Smart Systems, AI and Cybersecurity (CSSAIC)
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- 3D pose estimation from 2D pose data has been extensively studied due to its importance in human-robot interfaces, computer graphics and virtual reality. The significance of this paper it that, for the first time, it exploits randomness in the data to regenerate qualitatively more realistic humanlike 3D motions. As a result of the recognition of this work by Prof. Kazou Ishii and Prof. Akaiwa, Ghidary was invited to give a lecturer on multimodal learning at Furukawa lab Kyushu institute of Technology, Japan.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -