3D reconstruction of "in-the-wild" faces in images and videos
- Submitting institution
-
Imperial College of Science, Technology and Medicine
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2155
- Type
- D - Journal article
- DOI
-
10.1109/TPAMI.2018.2832138
- Title of journal
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Article number
- -
- First page
- 2638
- Volume
- 40
- Issue
- 11
- ISSN
- 0162-8828
- Open access status
- Compliant
- Month of publication
- May
- Year of publication
- 2018
- URL
-
-
- Supplementary information
-
10.1109/TPAMI.2018.2832138
- 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
-
6
- Research group(s)
-
-
- Citation count
- 14
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The paper presents the first method that fits a 3D Morphable Model (3DMM) of human faces in images and videos captured in unconstrained conditions. Our methodology outperforms the state-of-the-art in 3D face reconstruction by large margins. The technology of this work underpins the Imperial College startup FaceSoft, (https://www.linkedin.com/company/facesoftltd), which generates synthetic images to train large-scale face analysis systems for face recognition, facial expression recognition, face detection, and landmark localization. FaceSoft was acquired by a global technology company in 2020 in an undisclosed deal. The paper extends our highly-cited CVPR 2017 spotlight paper (8%/783, https://doi.org/10.1109/CVPR.2017.580).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
