Equidistant prototypes embedding for single sample based face recognition with generic learning and incremental learning

We develop a parameter-free face recognition algorithm which is insensitive to large variations in lighting, expression, occlusion, and age using a single gallery sample per subject. We take advantage of the observation that equidistant prototypes embedding is an optimal embedding that maximizes the...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Pattern recognition Jg. 47; H. 12; S. 3738 - 3749
Hauptverfasser: Deng, Weihong, Hu, Jiani, Zhou, Xiuzhuang, Guo, Jun
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Kidlington Elsevier Ltd 01.12.2014
Elsevier
Schlagworte:
Applied sciences
Detection, estimation, filtering, equalization, prediction
Exact sciences and technology
Face recognition
Feature extraction
Generic learning
Image processing
Information, signal and communications theory
Linear regression
One sample problem
Pattern recognition
Signal and communications theory
Signal processing
Signal, noise
Telecommunications and information theory
One sample problem
Feature extraction
Generic learning
Face recognition
Linear regression
Biometrics
Performance evaluation
State of the art
Prototype
Image processing
Updating
Mapping
Regression analysis
Pattern recognition
Algorithm
Learning
Least squares method
Database
Signal processing
Localization
Automatic recognition
ISSN:0031-3203, 1873-5142
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We develop a parameter-free face recognition algorithm which is insensitive to large variations in lighting, expression, occlusion, and age using a single gallery sample per subject. We take advantage of the observation that equidistant prototypes embedding is an optimal embedding that maximizes the minimum one-against-the-rest margin between the classes. Rather than preserving the global or the local structure of the training data, our method, called linear regression analysis (LRA), applies a least-square regression technique to map gallery samples to the equally distant locations, regardless of the true structure of training data. Further, a novel generic learning method, which maps the intraclass facial differences of the generic faces to the zero vectors, is incorporated to enhance the generalization capability of LRA. Using this novel method, learning based on only a handful of generic classes can largely improve the face recognition performance, even when the generic data are collected from a different database and camera set-up. The incremental learning based on the Greville algorithm makes the mapping matrix efficiently updated from the newly coming gallery classes, training samples, or generic variations. Although it is fairly simple and parameter-free, LRA, combined with commonly used local descriptors, such as Gabor representation and local binary patterns, outperforms the state-of-the-art methods for several standard experiments on the Extended Yale B, CMU PIE, AR, and FERET databases. •Equidistant prototypes embedding that maximizes the minimum one-against-the-rest margin between the classes.•A parameter-free model for the practical usage on one sample problem.•Effective generic learning based on the training samples collected from a different database.•An incremental learning algorithm that makes the model efficiently updated.•Extensive experimental results on the Extended Yale B, CMU PIE, AR, and FERET databases.
ISSN:0031-3203
1873-5142
DOI:10.1016/j.patcog.2014.06.020