Self-adaptive Feature Fusion Method for Improving LBP for Face Identification

Research output: Contribution to conferencePaper

Abstract

In a recent paper, a multi-scale information fusion method was presented to improve LBP for face identification. However, the additional parameters employed in that method cannot be automatically optimised. In this paper, a novel self-adaptive feature fusion method is proposed which extends the mLBP method by removing the need to optimise these parameters. Our method involves four steps. Firstly, a large number of initial features are generated. Then, we proposed a Fisher criteria-based method for evaluating the discriminative capabilities of different feature groups. After that, we proposed a model based on prism volume for selecting the optimal parameter combination. Finally, the resulting multi-scale feature are fused by a extended Euclidean distance fusion. Extensive experiments on two face databases have shown the proposed self-adaptive feature fusion method can find parameters that are optimal to the data in question, and can produce excellent classification performance.
LanguageEnglish
Pages373-383
DOIs
Publication statusE-pub ahead of print - 11 Oct 2017

Fingerprint

Information fusion
Prisms
Experiments

Keywords

  • Self-adaptive
  • Feature fusion
  • Face identification
  • LBP

Cite this

@conference{547a2ae3ed804acbab5524d6fd43736c,
title = "Self-adaptive Feature Fusion Method for Improving LBP for Face Identification",
abstract = "In a recent paper, a multi-scale information fusion method was presented to improve LBP for face identification. However, the additional parameters employed in that method cannot be automatically optimised. In this paper, a novel self-adaptive feature fusion method is proposed which extends the mLBP method by removing the need to optimise these parameters. Our method involves four steps. Firstly, a large number of initial features are generated. Then, we proposed a Fisher criteria-based method for evaluating the discriminative capabilities of different feature groups. After that, we proposed a model based on prism volume for selecting the optimal parameter combination. Finally, the resulting multi-scale feature are fused by a extended Euclidean distance fusion. Extensive experiments on two face databases have shown the proposed self-adaptive feature fusion method can find parameters that are optimal to the data in question, and can produce excellent classification performance.",
keywords = "Self-adaptive, Feature fusion, Face identification, LBP",
author = "Xin Wei and Hui Wang and Huan Wan and Bryan Scotney",
note = "Online ISBN 978-3-319-68345-4",
year = "2017",
month = "10",
day = "11",
doi = "10.1007/978-3-319-68345-4_33",
language = "English",
pages = "373--383",

}

TY - CONF

T1 - Self-adaptive Feature Fusion Method for Improving LBP for Face Identification

AU - Wei, Xin

AU - Wang, Hui

AU - Wan, Huan

AU - Scotney, Bryan

N1 - Online ISBN 978-3-319-68345-4

PY - 2017/10/11

Y1 - 2017/10/11

N2 - In a recent paper, a multi-scale information fusion method was presented to improve LBP for face identification. However, the additional parameters employed in that method cannot be automatically optimised. In this paper, a novel self-adaptive feature fusion method is proposed which extends the mLBP method by removing the need to optimise these parameters. Our method involves four steps. Firstly, a large number of initial features are generated. Then, we proposed a Fisher criteria-based method for evaluating the discriminative capabilities of different feature groups. After that, we proposed a model based on prism volume for selecting the optimal parameter combination. Finally, the resulting multi-scale feature are fused by a extended Euclidean distance fusion. Extensive experiments on two face databases have shown the proposed self-adaptive feature fusion method can find parameters that are optimal to the data in question, and can produce excellent classification performance.

AB - In a recent paper, a multi-scale information fusion method was presented to improve LBP for face identification. However, the additional parameters employed in that method cannot be automatically optimised. In this paper, a novel self-adaptive feature fusion method is proposed which extends the mLBP method by removing the need to optimise these parameters. Our method involves four steps. Firstly, a large number of initial features are generated. Then, we proposed a Fisher criteria-based method for evaluating the discriminative capabilities of different feature groups. After that, we proposed a model based on prism volume for selecting the optimal parameter combination. Finally, the resulting multi-scale feature are fused by a extended Euclidean distance fusion. Extensive experiments on two face databases have shown the proposed self-adaptive feature fusion method can find parameters that are optimal to the data in question, and can produce excellent classification performance.

KW - Self-adaptive

KW - Feature fusion

KW - Face identification

KW - LBP

UR - https://link.springer.com/chapter/10.1007/978-3-319-68345-4_33

U2 - 10.1007/978-3-319-68345-4_33

DO - 10.1007/978-3-319-68345-4_33

M3 - Paper

SP - 373

EP - 383

ER -