Mixture of Probabilistic Principal Component Analyzers for Shapes from Point Sets
- Submitting institution
-
The University of Leeds
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- UOA11-3955
- Type
- D - Journal article
- DOI
-
10.1109/TPAMI.2017.2700276
- Title of journal
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Article number
- -
- First page
- 891
- Volume
- 40
- Issue
- 4
- ISSN
- 0162-8828
- Open access status
- Technical exception
- Month of publication
- May
- 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
-
4
- Research group(s)
-
B - AI (Artificial Intelligence)
- Citation count
- 4
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper proposes a theoretical framework for statistical shape analysis represented by point clouds. The significance of the work relies in the rigour of the proposed variational Bayesian approach to address long standing challenges in the context of shape analysis: lack of point correspondences, unsuitability of conventional PCA based models when applied to shapes from distinct groups, and heuristic selections of number of PCA modes, clusters, etc. Given a population of shapes, the method fits a probability density function in the form of mixture of probabilistic PCAs. It has been applied to various organ shapes e.g. vertebrae and hearts.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
