An improvement to linear regression classification for face recognition
- PDF / 3,362,911 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 17 Downloads / 243 Views
ORIGINAL ARTICLE
An improvement to linear regression classification for face recognition Yali Peng1,2 · Jingcheng Ke1,2,3 · Shigang Liu1,2,3 · Jun Li4 · Tao Lei5 Received: 7 June 2017 / Accepted: 4 August 2018 © Springer-Verlag GmbH Germany, part of Springer Nature 2018
Abstract Linear regression classification (LRC) has attracted a great amount of attention owning to its promising performance in face recognition. However, its performance will dramatically decline in the scenario of limited training samples per class, particularly when only single training sample is available for a specific person. In this paper, a novel LRC based method is proposed to solve this problem. Specifically, we first perform LRC on the training set to obtain a kind of residual for each class. Next, a reverse representation residual is derived for each training sample of a class by exploiting the linear combination of the training samples of its nearest classes and the test sample. Then, we combine the reverse representation residuals of the class by an adaptive weighted-average approach to produce the other kind of residual. Finally, two kinds of residuals are fused to classify the test sample. Experimental results on the ORL, FERET, Libor94 and CMU-PIE face databases demonstrate that the proposed method obtains a higher recognition rate than some state-of-the-art face recognition methods. Keywords Single sample · Linear regression classification · Adaptive weighted-average approach
1 Introduction Despite the considerable progress made in the past few years, face recognition is still an open problem with a wide range of challenges in real-world application, such as various poses, * Shigang Liu [email protected] Yali Peng [email protected] Jingcheng Ke [email protected] Jun Li [email protected] Tao Lei [email protected] 1
Key Laboratory of Modern Teaching Technology, Ministry of Education, Xi’an 710062, China
2
Engineering Laboratory of Teaching Information Technology of Shaanxi Province, Xi’an 710119, Shaanxi, China
3
School of Computer Science, Shaanxi Normal University, Xi’an 710119, China
4
School of Automation, Southeast University Nanjing, Nanjing 210096, China
5
College of Electronics and Information Engineering, Shaanxi University of Science and Technology, Xi’an 710021, China
facial expressions and illuminations variances. In order to overcome these challenges and obtain a higher recognition rate, massive attempts have been made to address this problem [1–5]. Representation-based methods [6–15] are in essence the recognition methods which use the linear representation contribution of the training samples and test samples to perform face recognition. Conventional representation-based methods include spare representation classification (SRC) [7], collaborative representation based classification CRC [9], LRC [11], two phase test sample sparse representation classification (TPTSSR) [12], etc. SRC can be viewed as a norm representation-based method since it uses l1-regularization to obtain the coefficient ve
Data Loading...
