Development of an Improved Convolutional Neural Network for an Automated Face Based University Attendance System

In order to improve the standardization and accuracy of business process management of laboratory department in hospitals, combined with convolutional neural networks (CNN) and face recognition technology, an association application system of laboratory face recognition and test-tube barcode is designed by inputting patient's face and blood test-tube barcode into the system for storage. When the patient logs into the system again, the system uses the patient's face to automatically search for a matching test-tube barcode to obtain the test results. The simulation results show that the system can accurately recognize the face and match the corresponding test-tube barcode, and the accuracy and ROC of face recognition are 0.85 and 0.94, respectively. In addition, when the patient's face is within 5 m from the system camera, the accuracy of face recognition can reach 100%. It can be seen that the system designed in this paper shows good performance.

Face recognition is a biological tool that may be used to recognize people by their distinctive facial features. Numerous applications, including security, enforcement agencies, and consumer electronics, have made use of this technology. Face detection and recognition system works by analyzing a person's individual facial features to a database of recognized faces. The system can recognize the person if there is a match. We employ a variety of already trained models, such as those for image processing, pattern recognition, and artificial neural network method (ANN). The accuracy of facial recognition has been considerably improved by the advent of deep learning technologies. CNNs (Convolutional Neural Networks) with deep learning have significantly improved picture categorization. The Computer Vision and Machine Learning Group's deep learning model, DeepID, has achieved recognition rate of 98% on the ORL database for attendance systems. Key Words: Face Recognition, Biometric, Image processing, Pattern Recognition, Artificial Neural Network

Convolutional neural network (CNN) have been widely used in face recognition. The illumination and the size of the data set affect the accuracy of face recognition based on CNN. In order to improve the accuracy of face recognition based on CNN, an improved face recognition algorithm based on CNN with extend local binary pattern (ELBP) and deep convolutional generative adversarial network (DCGAN) is proposed. ELBP uses a circular operator which can be of any size, and the coverage area can be adjusted arbitrarily. It has gray and rotation invariance, and has good robustness to illumination. In real scenes, it is more difficult to obtain face photos of a person with different lighting and different scenes. In order to solve face recognition in the case of small data sets, this paper uses DCGAN to generate new face pictures based on the original pictures. By expanding the data set, the accuracy of face recognition is improved. Experiments have proved that the proposed method in this paper has higher accuracy in face recognition than traditional methods.

Face recognition is a computationally challenging classification task. Deep convolutional neural networks (DCNNs) are brain-inspired algorithms that have recently reached human-level performance in face and object recognition. However, it is not clear to what extent DCNNs generate a human-like representation of face identity. We have recently revealed a subset of facial features that are used by humans for face recognition. This enables us now to ask whether DCNNs rely on the same facial information and whether this human-like representation depends on a system that is optimized for face identification. In the current study, we examined the representation of DCNNs of faces that differ in features that are critical or non-critical for human face recognition. Our findings show that DCNNs optimized for face identification are tuned to the same facial features used by humans for face recognition. Sensitivity to these features was highly correlated with performance of the DCNN on a benchmark face recognition task. Moreover, sensitivity to these features and a view-invariant face representation emerged at higher layers of a DCNN optimized for face recognition but not for object recognition. This finding parallels the division to a face and an object system in high-level visual cortex. Taken together, these findings validate human perceptual models of face recognition, enable us to use DCNNs to test predictions about human face and object recognition as well as contribute to the interpretability of DCNNs.

Deep learning, in particular Convolutional Neural Network (CNN), has achieved promising results in face recognition recently. However, it remains an open question: why CNNs work well and how to design a 'good' architecture. The existing works tend to focus on reporting CNN architectures that work well for face recognition rather than investigate the reason. In this work, we conduct an extensive evaluation of CNN-based face recognition systems (CNN-FRS) on a common ground to make our work easily reproducible. Specifically, we use public database LFW (Labeled Faces in the Wild) to train CNNs, unlike most existing CNNs trained on private databases. We propose three CNN architectures which are the first reported architectures trained using LFW data. This paper quantitatively compares the architectures of CNNs and evaluate the effect of different implementation choices. We identify several useful properties of CNN-FRS. For instance, the dimensionality of the learned features can be significantly reduced without adverse effect on face recognition accuracy. In addition, traditional metric learning method exploiting CNN-learned features is evaluated. Experiments show two crucial factors to good CNN-FRS performance are the fusion of multiple CNNs and metric learning. To make our work reproducible, source code and models will be made publicly available.

In the real environment, face recognition based on video or image files is easily affected by the acquisition angle, expression, illumination and other conditions, and general machine learning and pattern recognition face recognition algorithms also have problems of accuracy and adaptability. Therefore, based on the current technical development, this paper takes the deep learning algorithm model as the core, uses MTCNN multi-task convolutional neural network with FaceNet model to complete the face detection and face recognition of video image files, and completes the corresponding data training with the help of Keras framework and TensorFlow class library to form an intelligent machine that can support the call of Web server. At the same time, ASP.NET framework and C# language are used to build the Web server, design and develop each functional module, improve the deployment of the corresponding API interface, build the front-end interactive interface, and form the online face recognition system. The overall design of the system adopts B/S architecture, which supports users to access the Web Server through simple request operation to complete the corresponding face detection and recognition of the input video image files, and return the comparison recognition results to the front page for display. The system will greatly improve the precision and accuracy of face recognition, and make a useful attempt to further expand the application scenarios of face recognition technology.

Improving face verification using facial marks and deep CNN: IARPA Janus benchmark-A

In order to improve the low accuracy of the face recognition methods in the case of e-health, this paper proposed a novel face recognition approach, which is based on convolutional neural network (CNN). In detail, through resolving the convolutional kernel, rectified linear unit (ReLU) activation function, dropout, and batch normalization, this novel approach reduces the number of parameters of the CNN model, improves the non-linearity of the CNN model, and alleviates overfitting of the CNN model. In these ways, the accuracy of face recognition is increased. In the experiments, the proposed approach is compared with principal component analysis (PCA) and support vector machine (SVM) on ORL, Cohn-Kanade, and extended Yale-B face recognition data set, and it proves that this approach is promising.

Optimization of face recognition algorithm based on deep learning multi feature fusion driven by big data

The facial recognition system is an application tool that uses artificial intelligence technology and biometrics technology to analyze and recognize the facial feature information of the human face. It is widely used in various fields, such as attendance and access control management in schools and companies, identity monitoring in stations and stores, facial recognition for fugitive criminals, and facial payment on mobile terminals. However, due to the short development time of the facial recognition system, the facial recognition system has the problem of low recognition accuracy when the recognized object is not cooperative. Although some scholars have proposed the region of interest (ROI)-K nearest neighbor algorithm (KNN) convolutional neural network theory by using the ROI and KNN and applied it to face recognition, the facial recognition system based on ROI-KNN convolutional neural network did not solve the problems of insufficient facial recognition accuracy and insufficient security. Under the conditions of insufficient illumination, excessive expression change, occlusion, high similarity of different individuals, and dynamic recognition, the recognition effect of the facial recognition system based on the ROI-KNN convolutional neural network is relatively limited. Therefore, to make the recognition accuracy of the facial recognition system higher and to make the facial recognition system play a greater role in the social and economic fields, this paper used the adaptive quantum genetic algorithm, the improved marker line graph genetic algorithm, and the feature weight value genetic algorithm to study the facial recognition system of the ROI-KNN convolutional neural network. The research results showed that after improving the ROI-KNN convolutional neural network based on the genetic algorithm, the recognition accuracy of the facial recognition system was increased by 4.99%, the recognition speed was increased by 7.46%, and the recognition security was increased by 2.66%.

Face detection and recognition plays an important role in many occasions. This study explored the application of convolutional neural network in face detection and recognition. Firstly, convolutional neural network was briefly analyzed, and then a face detection model including three convolution layers, four pooling layers, introduction layers and three fully connected layers was designed. In face recognition, the self-learning convolutional neural network (CNN) model for global and local extended learning and Spatial Pyramid Pooling (SPP)-NET model were established. LFW data sets were used as model test samples. The results showed that the face detection model had an accuracy rate of 99%. In face recognition, the self-learning CNN model had an accuracy rate of 94.9% accuracy, and the SPP-Net model had an accuracy rate of 92.85%. It suggests that the face detection and recognition model based on convolutional neural network has good accuracy, and the face recognition efficiency of self-learning CNN model was better, which deserves further research and promotion.

An empirical study of the impact of masks on face recognition

Objective:: The purpose of this paper was to use Machine Learning (ML) techniques to extract facial features from images. Accurate face detection and recognition has long been a problem in computer vision. According to a recent study, Local Binary Pattern (LBP) is a superior facial descriptor for face recognition. A person's face may make their identity, feelings, and ideas more obvious. In the modern world, everyone wants to feel secure from unauthorized authentication. Face detection and recognition help increase security; however, the most difficult challenge is to accurately recognise faces without creating any false identities. Methods:: The proposed method uses a Local Binary Pattern Histogram (LBPH) and Convolution Neural Network (CNN) to preprocess face images with equalized histograms. Results:: LBPH in the proposed technique is used to extract and join the histogram values into a single vector. The technique has been found to result in a reduction in training loss and an increase in validation accuracy of over 96.5%. Prior algorithms have been reported with lower accuracy when compared to LBPH using CNN. Conclusion:: This study demonstrates how studying characteristics produces more precise results, as the number of epochs increases. By comparing facial similarities, the vector has generated the best result.

In recent years, designing new effective Deep Convolutional Neural Network (DCNN) architectures and loss functions are two crucial trends in improving face recognition (FR) accuracy. However, building an optimal DCNN and ameliorating FR performance are still the main challenges for researchers. Thus, we first investigate and analyzes the effect of several novel effective loss functions based on softmax on DCNN with the ResNet architecture. We then propose an ensemble learning, namely EnsFace, by taking advantage of recent novel FR methods based on CosFace, ArcFace, and MagFace. EnsFace elaborates on the voting mechanism that utilizes non-optimal pre-trained models to obtain better discriminative ability for FR. To prove the performance of both the speed and accuracy of EnsFace, we carry out rigorous experiments using several popular benchmarks, including LFW, CFP-FP, and AgeDB-30, as well as two renovations of LFW: CALFW and CPLFW. The results of our experiments achieve state-of-the-art figures, which show the proposed method’s massive potential to improve FR performance. Ablation studies and overall benchmarks undeniably prove the effectiveness of our EnsFace.

with rapid development of deep learning technology, face recognition based on deep convolutional neural network becomes one of the main research methods. In order to solve the problems of information loss and equal treatment of each element in the input feature graph in the traditional pooling method of convolutional neural network, a face recognition algorithm based on convolutional neural network is proposed in this paper. First, MTCNN algorithm is used to detect the faces and do gray processing, and then a local weighted average pooling method based on local concern strategy is designed and a convolutional neural network based on VGG16 to recognize faces is constructed which is finally compared with common convolutional neural network. The experimental results show that this method has good face recognition accuracy in common face databases.