For the past few decades, biometric Face Recognition (FR) is the active research area in different domains such as image processing, pattern recognition, etc. The existing FR system has several limitations such as Single Sample Problem, maximum Reconstruction Errors (REs), these problems decreases the FR rate. In this research paper, an efficient FR method is proposed, namely Grey Wolf Optimizer based Linear Collaborative Discriminant Regression Classification (GWO-LCDRC). The optimization technique of GWO algorithm is applied in LCDRC to select the relevant weight value in LCDRC. For every training sample, optimal weight values are selected to improve the recognition rate. The proposed GWO-LCDRC method maximize the collaboration of Between Class RE (BCRE) and minimize the Within Class RE (WCRE). An experimental analysis conducted on the two standard facial databases namely ORL and YALE. The proposed GWO-LCDRC method improved the performance of recognition accuracy with respect to different training samples. Also, the performance is compared with the traditional methods Linear Regression Classification (LRC), Linear Discriminant Regression Classification (LDRC), and LCDRC. The overall experiment demonstrated that the proposed GWO-LCDRC method achieved approximately 3% and 6.5% of FR accuracy improvement with respect to ORL and YALE respectively.

No abstract

Face recognition is one of the important biometric authentication research areas for security purposes in many fields such as pattern recognition and image processing. However, the human face recognitions have the major problem in machine learning and deep learning techniques, since input images vary with poses of people, different lighting conditions, various expressions, ages as well as illumination conditions and it makes the face recognition process poor in accuracy. In the present research, the resolution of the image patches is reduced by the max pooling layer in convolutional neural network (CNN) and also used to make the model robust than other traditional feature extraction technique called local multiple pattern (LMP). The extracted features are fed into the linear collaborative discriminant regression classification (LCDRC) for final face recognition. Due to optimization using CNN in LCDRC, the distance ratio between the classes has maximized and the distance of the features inside the class reduces. The results stated that the CNN-LCDRC achieved 93.10% and 87.60% of mean recognition accuracy, where traditional LCDRC achieved 83.35% and 77.70% of mean recognition accuracy on ORL and YALE databases respectively for the training number 8 (i.e. 80% of training and 20% of testing data).

No abstract