Recognition Rate For Face Images Research Articles

A Robust Face Recognition Algorithm Based on an Improved Generative Confrontation Network

Objective: In practical applications, an image of a face is often partially occluded, which decreases the recognition rate and the robustness. Therefore, in response to this situation, an effective face recognition model based on an improved generative adversarial network (GAN) is proposed. Methods: First, we use a generator composed of an autoencoder and the adversarial learning of two discriminators (local discriminator and global discriminator) to fill and repair an occluded face image. On this basis, the Resnet-50 network is used to perform image restoration on the face. In our recognition framework, we introduce a classification loss function that can quantify the distance between classes. The image generated by the generator can only capture the rough shape of the missing facial components or generate the wrong pixels. To obtain a clearer and more realistic image, this paper uses two discriminators (local discriminator and global discriminator, as mentioned above). The images generated by the proposed method are coherent and minimally influence facial expression recognition. Through experiments, facial images with different occlusion conditions are compared before and after the facial expressions are filled, and the recognition rates of different algorithms are compared. Results: The images generated by the method in this paper are truly coherent and have little impact on facial expression recognition. When the occlusion area is less than 50%, the overall recognition rate of the model is above 80%, which is close to the recognition rate pertaining to the non-occluded images. Conclusions: The experimental results show that the method in this paper has a better restoration effect and higher recognition rate for face images of different occlusion types and regions. Furthermore, it can be used for face recognition in a daily occlusion environment, and achieve a better recognition effect.

A new Uyghur face recognition algorithm

In this paper, in light of the reduced recognition rate and poor robustness of Uyghur face images under non-uniform illumination, a Histogram of Oriented Gradients (HOG) algorithm combining Discrete Cosine Transform (DCT) with Laplace filter was proposed. Firstly, the Uyghur face images were processed by Laplace filter to highlight their edges and texture features and reduce the interference of illumination; secondly, the Uyghur face images processed by Laplace filter were transformed by DCT, which can effectively filter the high frequency information and retain the low frequency information resulted from the DCT; thirdly, the Uyghur face images were reconstructed by Inverse Discrete Cosine Transform (IDCT), and the Uyghur face images reconstructed by IDCT were still similar to the original Uyghur face images, so that the non-low frequency information in the Uyghur face images was removed and the dimension of the Uyghur face images was reduced; finally, the inherent features of the final Uyghur face images were extracted using Histogram of Oriented Gradients (HOG) operator, and the final Uyghur face images were classified using the Nearest Neighbor Method. The simulation experiment results showed that the proposed algorithm further improved the recognition rate of the face images in face database — as high as 95% in the Yale B face database under different feature dimensions, and 98.5% in the Uyghur face database built by the members of the research group, respectively, which was superior to other traditional algorithms, and had strong robustness and good real time performance.

A New Hybrid Approach Using PCA for Pose Invariant Face Recognition

In this paper a new hybrid approach using PCA for pose invariant face recognition is proposed. In this proposed approach three algorithms are combined to make a new hybrid approach. The first step is to detect face and its part. It is done by well known Viola Jones algorithm. In this proposed new hybrid approach using PCA, five parts of face image are detected and these are face, left eye, right eye, nose and mouth. The second step is to find local binary pattern (LBP) of each part. LBP extracts features from the detected faces and its parts. The third step is to apply PCA on each extracted feature for recognition. It is seen from the experimental results that proposed hybrid approach using PCA gives improved recognition rate for face images with different facial expression and poses. It is when compared with conventional PCA, PCA+ Wavelet, 2DPCA, 2DPCA + DWT and LBP algorithms shows improved recognition rate for face images. The accuracy of the conventional PCA and hybrid approach using PCA are evaluated under the conditions of varying expression and pose. The images are taken from ORL face databases.

An Improved Face Recognition Using Neighborhood Defined Modular Phase Congruency Based Kernel PCA

A face recognition algorithm based on NMPKPCA algorithm presented in this paper. The proposed algorithm when compared with conventional Principal component analysis (PCA) algorithms has an improved recognition Rate for face images with large variations in illumination, facial expressions. In this technique, first phase congruency features are extracted from the face image so that effects due to illumination variations are avoided by considering phase component of image. Then, face images are divided into small sub images and the kernel PCA approach is applied to each of these sub images. but, dividing into small or large modules creates some problems in recognition. So a special modulation called neighborhood defined modularization approach presented in this paper, so that effects due to facial variations are avoided. Then, kernel PCA has been applied to each module to extract features. So a feature extraction technique for improving recognition accuracy of a visual image based facial recognition system presented in this paper.

An improved face recognition technique based on modular PCA approach

A face recognition algorithm based on modular PCA approach is presented in this paper. The proposed algorithm when compared with conventional PCA algorithm has an improved recognition rate for face images with large variations in lighting direction and facial expression. In the proposed technique, the face images are divided into smaller sub-images and the PCA approach is applied to each of these sub-images. Since some of the local facial features of an individual do not vary even when the pose, lighting direction and facial expression vary, we expect the proposed method to be able to cope with these variations. The accuracy of the conventional PCA method and modular PCA method are evaluated under the conditions of varying expression, illumination and pose using standard face databases.