A Face Recognition Algorithm Based on Intermediate Layers Connected by the CNN

2021 ◽  
Author(s):  
Yuxiang Long
Keyword(s):  
Feature Extraction ◽  
Face Recognition ◽  
Network Model ◽  
Recognition Algorithm ◽  
Image Features ◽  
Face Image ◽  
Face Features ◽  
Intermediate Layers ◽  
Training Samples ◽  
The Face

Face recognition is difficult due to the higher dimension of face image features and fewer training samples. Firstly, in order to improve the performance of feature extraction, we inventively construct a double hierarchical network structure convolution neural network (CNN) model. The front-end network adopts a relatively simple network model to achieve rough feature extraction from input images and obtain multiple suspect face candidate windows. The back-end network uses a relatively complex network model to filter the best detection window and return the face size and position by nonmaximum suppression. Then, in order to fully extract the face features in the optimal window, a face recognition algorithm based on intermediate layers connected by the deep CNN is proposed in this paper. Based on AlexNet, the front, intermediate and end convolution layers are combined by deep connection. Then, the feature vector describing the face image is obtained by the operation of the pooling layer and the full connection layer. Finally, the auxiliary classifier training method is used to train the model to ensure the effectiveness of the features of the intermediate layer. Experimental results based on open face database show that the recognition accuracy of the proposed algorithm is higher than that of other face recognition algorithms compared in this paper.

The Face Detection Research Based on Multi-Scale and Rectangle Feature

2012 ◽  
Vol 198-199 ◽  
pp. 1383-1388
Author(s):  
Hong Hai Liu ◽  
Xiang Hua Hou
Keyword(s):  
Feature Extraction ◽  
Image Features ◽  
Face Image ◽  
High Dimensions ◽  
Feature Extraction Method ◽  
Face Features ◽  
Haar Feature ◽  
The Face ◽  
Speed Up ◽  
Feature Expression

When extracting the face image features based on pixel distribution in face image, there always exist large amount of calculation and high dimensions of feature sector generated after feature extraction. This paper puts forward a feature extraction method based on prior knowledge of face and Haar feature. Firstly, the Haar feature expressions of face images are classified and the face features are decomposed into edge feature, line feature and center-surround feature, which are further concluded into the expressions of two rectangles, three rectangles and four rectangles. In addition, each rectangle varies in size. However, for face image combination, there exist too much redundancy and large calculation amount in this kind of expression. In order to solve the problem of large amount of calculation, the integral image is adopted to speed up the rectangle feature calculation. In addition, the thought based on classified trainer is adopted to reduce the redundancy expression. The results show that using face image of Haar feature expression can improve the speed and efficiency of recognition.

scholarly journals Unrestricted Face Recognition Algorithm Based on Transfer Learning on Self-Pickup Cabinet

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Zhixue Liang
Keyword(s):  
Face Recognition ◽  
Transfer Learning ◽  
Recognition Rate ◽  
Recognition Algorithm ◽  
Face Image ◽  
Delivery Device ◽  
Face Images ◽  
Region Extraction ◽  
The Face ◽  
Enhancement Algorithm

In the contactless delivery scenario, the self-pickup cabinet is an important terminal delivery device, and face recognition is one of the efficient ways to achieve contactless access express delivery. In order to effectively recognize face images under unrestricted environments, an unrestricted face recognition algorithm based on transfer learning is proposed in this study. First, the region extraction network of the faster RCNN algorithm is improved to improve the recognition speed of the algorithm. Then, the first transfer learning is applied between the large ImageNet dataset and the face image dataset under restricted conditions. The second transfer learning is applied between face image under restricted conditions and unrestricted face image datasets. Finally, the unrestricted face image is processed by the image enhancement algorithm to increase its similarity with the restricted face image, so that the second transfer learning can be carried out effectively. Experimental results show that the proposed algorithm has better recognition rate and recognition speed on the CASIA-WebFace dataset, FLW dataset, and MegaFace dataset.

Face Recognition Algorithm Based on 3D Point Cloud Acquired by Mixed Image Sensor

2019 ◽  
Vol 23 (2) ◽  
pp. 366-369
Author(s):  
Tang-Tang Yi ◽  
Keyword(s):  
Face Recognition ◽  
Point Cloud ◽  
Recognition Accuracy ◽  
Image Sensor ◽  
Recognition Algorithm ◽  
Face Image ◽  
3D Point Cloud ◽  
3D Face ◽  
Face Images ◽  
The Face

In order to solve the problem of low recognition accuracy in recognition of 3D face images collected by traditional sensors, a face recognition algorithm for 3D point cloud collected by mixed image sensors is proposed. The algorithm first uses the 3D wheelbase to expand the face image edge. According to the 3D wheelbase, the noise of extended image is detected, and median filtering is used to eliminate the detected noise. Secondly, the priority of the boundary pixels to recognize the face image in the denoising image recognition process is determined, and the key parts such as the illuminance line are analyzed, so that the recognition of the 3D point cloud face image is completed. Experiments show that the proposed algorithm improves the recognition accuracy of 3D face images, which recognition time is lower than that of the traditional algorithm by about 4 times, and the recognition efficiency is high.

Face Recognition with Single Sample Based on Candide-3 Reconstruction Model

2013 ◽  
Vol 380-384 ◽  
pp. 3623-3628 ◽  
Author(s):  
Nan Deng ◽  
Ya Bo Pei ◽  
Zheng Guang Xu
Keyword(s):  
Support Vector Machine ◽  
Feature Extraction ◽  
Face Recognition ◽  
Recognition Rate ◽  
Single Sample ◽  
Support Vector ◽  
Training Samples ◽  
Virtual Images ◽  
The Face ◽  
Reconstruction Model

In this study, we present a method for virtual images generation based on Candide-3 model to increase the number of training samples for the face recognition with single sample, where the Principle Component Analysis is used for feature extraction and the test samples are classified by the method of Support Vector Machine (SVM). Experimental results on from the YaleB and ORL databases show that the recognition rate of the face recognition with single sample can be improved by the proposed method.

scholarly journals Face Recognition Based on Histogram Equalization and LBP Algorithm

2019 ◽  
Vol 8 (3) ◽  
pp. 33
Author(s):  
Herman Kh. Omar ◽  
Nada E. Tawfiq
Keyword(s):  
Face Recognition ◽  
Texture Classification ◽  
Histogram Equalization ◽  
Recognition System ◽  
Input Image ◽  
Image Features ◽  
Face Image ◽  
Wide Field ◽  
Face Recognition System ◽  
The Face

In the recent time bioinformatics take wide field in image processing. Face recognition which is basically the task of recognizing a person based on its facial image. It has become very popular in the last two decades, mainly because of the new methods developed and the high quality of the current visual instruments. There are different types of face recognition algorithms, and each method has a different approach to extract the image features and perform the matching with the input image. In this paper the Local Binary Patterns (LBP) was used, which is a particular case of the Texture Spectrum model, and powerful feature for texture classification. The face recognition system consists of recognizing the faces acquisition from a given data base via two phases. The most useful and unique features of the face image are extracted in the feature extraction phase. In the classification the face image is compared with the images from the database. The proposed algorithm for face recognition in this paper adopt the LBP features encode local texture information with default values. Apply histogram equalization and Resize the image into 80x60, divide it to five blocks, then Save every LBP feature as a vector table. Matlab R2019a was used to build the face recognition system. The Results which obtained are accurate and they are 98.8% overall (500 face image).

scholarly journals A Face Recognition System using Directional Binary Code Algorithm and Multi-SVM

2019 ◽  
Vol 8 (6S3) ◽  
pp. 1201-1208
Keyword(s):  
Feature Extraction ◽  
Face Recognition ◽  
Binary Code ◽  
Recognition System ◽  
Extraction Methods ◽  
Recognition Algorithm ◽  
Svm Classifier ◽  
Feature Extraction Method ◽  
Face Recognition System ◽  
The Face

Over past few years, face recognition technology plays an important function in the development of biometric identifier with less time consuming and computational overhead. Many researchers were put their effort to develop face recognition algorithm involves three distinct steps such as detection, unique faceprint creation and finally verification. Traditional Local binary pattern based face recognition system slow down the recognition speed, high computational complexity and does not give the directional data of the picture. In order to overcome the above limitation, a novel face recognition system is proposed by employing the advantage of Directional Binary Code (DBC) feature extraction method. The face images features are extracted from DBC are generally smoother than other feature extraction methods. The images with blur creation, pose changes, and illumination is applied and stored in the database. For blur creation various filters such as Average filter, Gaussian filter and Motion filter are used. By using Directional Binary Code method, the face is detected and extracted. Then the same algorithm is used for input images and with help of Multi-SVM classifier multiple images in the database is compared and shows the matched images. Finally, simulation result shows the implemented results in term of its recognition speed and computation complexity.

scholarly journals A Curvelet Domain Face Recognition Scheme Based on Local Dominant Feature Extraction

2012 ◽  
Vol 2012 ◽  
pp. 1-13 ◽  
Author(s):  
Hafiz Imtiaz ◽  
Shaikh Anowarul Fattah
Keyword(s):  
Feature Extraction ◽  
Face Recognition ◽  
Selection Criterion ◽  
Curvelet Transform ◽  
Principal Component ◽  
Face Image ◽  
Spatial Variations ◽  
Threshold Criterion ◽  
Dominant Feature ◽  
The Face

A feature extraction algorithm is introduced for face recognition, which efficiently exploits the local spatial variations in a face image utilizing curvelet transform. Although multi-resolution ideas have been profusely employed for addressing face recognition problems, theoretical studies indicate that digital curvelet transform is an even better method due to its directional properties. Instead of considering the entire face image, an entropy-based local band selection criterion is developed for feature extraction, which selects high-informative horizontal bands from the face image. These bands are segmented into several small spatial modules to capture the local spatial variations precisely. The effect of modularization in terms of the entropy content of the face images has been investigated. Dominant curvelet transform coefficients corresponding to each local region residing inside the horizontal bands are selected, based on the proposed threshold criterion, as features, which not only drastically reduces the feature dimension but also provides high within-class compactness and high between-class separability. A principal component analysis is performed to further reduce the dimensionality of the feature space. Extensive experimentation is carried out upon standard face databases and a very high degree of recognition accuracy is achieved even with a simple Euclidean distance based classifier.

A Novel Mechanism of Face Recognition Using Stepwise Linear Discriminant Analysis and Linear Vector Quantization Classifiers

2017 ◽  
Vol 7 (7) ◽  
pp. 48
Author(s):  
Abdul Quyoom
Keyword(s):  
Face Recognition ◽  
Discriminant Analysis ◽  
Linear Discriminant Analysis ◽  
Vector Quantization ◽  
Recognition System ◽  
Face Image ◽  
Linear Discriminant ◽  
Linear Vector ◽  
Face Features ◽  
The Face

Face recognition is a hard and special case of computer vision and pattern recognition. It is a challenging problem due to various kinds of variations of face images.  This paper proposes a robust face recognition system. Here stepwise linear discriminant analysis (SWLDA) is used for the feature extraction and Linear Vector Quantization (LVQ) Classifier is used for face recognition. The main focus of SWLDA is to select localized features from the face. In order to increase the low-between-class variance and to reduce within-class-variance among different expression classes and use F-test value through which results are analyzed. In recognition, firstly face is detected using canny edge detection method, after face detection SWLDA is employed to extract the face features, and end linear vector quantization is applied for face recognition. To achieve optimum results and increase the robustness of the proposed system, experiments are performed on various different samples of face image, which consist of face image with the different pose and facial expression in order to validate the system, we use two famous datasets which include Yale and ORL face database.

DISCRIMINATORY SPARSE CODING AND ITS APPLICATION TO FACE RECOGNITION

2003 ◽  
Vol 03 (03) ◽  
pp. 503-521
Author(s):  
JUN SUN ◽  
WENYUAN WANG ◽  
QING ZHUO ◽  
CHENGYUAN MA
Keyword(s):  
Feature Extraction ◽  
Sparse Coding ◽  
Recognition Performance ◽  
Classification Performance ◽  
Image Features ◽  
Face Image ◽  
Image Feature ◽  
Training Samples ◽  
The Subject ◽  
Independent Features

Feature extraction is very important in the subject of pattern recognition. Sparse coding is an approach for extracting the independent features of an image. The image features extracted by sparse coding have led to better recognition performance as compared to those from traditional PCA-based methods. A new discriminatory sparse coding (DSC) algorithm is proposed in this paper to further improve the classification performance. Based on reinforcement learning, DSC encodes the training samples by individual class rather than by individual image as in sparse coding. Having done that it will produce a set of features with large and small intraclass variations, which is very suitable for recognition tasks. Experiments are performed on face image feature extraction and recognition. Compared with the traditional PCA- and ICA-based methods, DSC shows a much better recognition performance.

Sign in / Sign up

Export Citation Format

Share Document