A face recognition framework based on a pool of techniques and differential evolution

Plichoski, Guilherme Felippe; Chidambaram, Chidambaram; Parpinelli, Rafael Stubs

doi:10.1016/j.ins.2020.06.054

Cited by 25 publications

(10 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evolve a current population, a DE algorithm uses crossover, selection, and mutation operators. A number of real world optimization problems in various fields have been successfully solved by DE algorithms, such as image processing [4], microwave engineering [5], signal processing [6], chemical engineering [7], artificial neural networks [8], bioinformatics [9], power systems [10], pattern recognition [11], robotics [12], convolutional neural network training [13], scheduling [14], communication [15].…”

Section: Introductionmentioning

confidence: 99%

Convergence Track Based Adaptive Differential Evolution Algorithm (CTbADE)

Abbas¹,

Malik²,

Saudagar³

et al. 2022

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

One of the challenging problems with evolutionary computing algorithms is to maintain the balance between exploration and exploitation capability in order to search global optima. A novel convergence track based adaptive differential evolution (CTbADE) algorithm is presented in this research paper. The crossover rate and mutation probability parameters in a differential evolution algorithm have a significant role in searching global optima. A more diverse population improves the global searching capability and helps to escape from the local optima problem. Tracking the convergence path over time helps enhance the searching speed of a differential evolution algorithm for varying problems. An adaptive powerful parameter-controlled sequences utilized learning period-based memory and following convergence track over time are introduced in this paper. The proposed algorithm will be helpful in maintaining the equilibrium between an algorithm's exploration and exploitation capability. A comprehensive test suite of standard benchmark problems with different natures, i.e., unimodal/multimodal and separable/non-separable, was used to test the convergence power of the proposed CTbADE algorithm. Experimental results show the significant performance of the CTbADE algorithm in terms of average fitness, solution quality, and convergence speed when compared with standard differential evolution algorithms and a few other commonly used state-of-the-art algorithms, such as jDE, CoDE, and EPSDE algorithms. This algorithm will prove to be a significant addition to the literature in order to solve real time problems and to optimize computational models with a high number of parameters to adjust during the problem-solving process.

show abstract

Section: Introductionmentioning

confidence: 99%

Convergence Track Based Adaptive Differential Evolution Algorithm (CTbADE)

Abbas¹,

Malik²,

Saudagar³

et al. 2022

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

“…Up to date, four main kinds of methods for SSPP face recognition are reported, including virtual sample generation [19][20][21], classifier [22][23][24], image partitioning [25][26][27][28][29] and generic learning [30][31][32][33][34][35][36][37][38][39][40]. The virtual sample generation method uses the singular value decomposition (SVD) [19], pose and illumination variability [20], the lower-upper decomposition [21] etc.…”

Section: Introductionmentioning

confidence: 99%

“…to make the distance between the same class of frontal gallery sample and the probe sample close to the distance between the different class. Plichoskia et al [22] proposed an SSPP method based on differential evolution algorithm, applied the differential evolution method to the face recognition framework for the first time, which has good performance under constrained environment. Pang et al [23] proposed a iterative dynamic generic learning (IDGL) method to handle SSPP problem with a contaminated biometric enrolment database, which has good performance under unconstrained environment.…”

Section: Introductionmentioning

confidence: 99%

“…What makes the situation even worse is that the probe samples are usually captured with video surveillance cameras under unconstrained conditions [8,9]. Therefore, SSPP face recognition in this unconstrained environment is very challenging task due to, first, the strong intraclass variations such as illumination [10][11][12], pose [12,13], facial expression [14][15][16] and occlusion [17,18], and second, the lack of gallery samples.…”

Section: Introductionmentioning

confidence: 99%

“…Plichoskia et al. [22] proposed an SSPP method based on differential evolution algorithm, applied the differential evolution method to the face recognition framework for the first time, which has good performance under constrained environment. Pang et al.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Single sample per person face recognition algorithm based on the robust prototype dictionary and robust variation dictionary construction

Xue

Ren

2021

IET Image Processing

View full text Add to dashboard Cite

Single sample per person (SSPP) face recognition uses only a single face image of each subject in the gallery set to recognize the probe sample. Since single sample cannot provide intra-class variation information, the matching accuracy of the gallery faces with the faces captured in unconstrained video is usually low. Recently, in order to improve the accuracy, the sparse representation-based classification (SRC) technology has been extended to generic learning method, which uses prototype and variation dictionary (P+V) model for face recognition. Because the inter-class scatter between atoms in prototype dictionary is not big enough, and the intra-class scatter in constructed variation dictionary (such as posture and expression) is not rich enough, the robustness of P+V model for SSPP face recognition is poor. To solve this problem, a robust prototype dictionary and robust variation dictionary construction (RPRV) method is proposed. First, a set of atoms is obtained by dictionary learning method using gallery images and generic images. Second, some effective atoms are selected by the proposed function index method. Finally, the robust prototype dictionary (RP) and the robust variation dictionary (RV) are represented linearly using these effective atoms, respectively. The face recognition is performed according to the proposed RP+RV model. Experiment results using public datasets show that the proposed RPRV has strong robustness for the face captured under the unconstrained environment. Comparison results show that the proposed RPRV method outperform state-of-the-art SRC-based methods for SSPP face recognition.

show abstract