Accuracy Of Face Detection Research Articles

Покращення моделей розпізнавання облич за допомогою навчання подібності, розкладу зміни темпу навчання та аугментацій

Розпізнавання облич — одна з основних задач комп’ютерного зору, актуальна в силу її практичної значущості та викликає велику зацікавленість широкого кола науковців. І хоча дослідження у сфері відбувались з початку розвитку комп’ютерного зору, адекватних результатів змогли досягнути лише за допомогою згорткових нейронних мереж. У даній роботі проведено порівняльний аналіз методів розпізнавання обличчя до згорткових нейронних мереж. Розглянуто метод навчання подібності, аугментації та розклади зміни темпу навчання. Проведено ряд експериментів, виконано порівняльний аналіз розглянутих методів покращення згорткових нейронних мереж, у результаті отримано універсальний алгоритм для навчання моделі розпізнавання облич. У роботі використано SE-ResNet50 як єдину мережу для експериментів. Навчання подібності — це метод, за допомогою якого можливо досягнути достатньої точності. Перенавчання є серйозною проблемою для нейронних мереж, зокрема тому, що мають дуже багато параметрів, а даних зазвичай не настільки багато, щоб гарантувати узагальнюючу можливість моделі. Додаткова розмітка даних може займати багато часу і бути доволі дорогою, тому виник такий підхід, як аугментації. Аугментації штучно збільшують тренувальний набір даних, тому цілком природньо, що метод аугментації у всіх експериментах покращив результати відносно початкового експерименту. Різні степені та більш агресивні форми аугментації у задачі розпізнавання облич у даній роботі приводив до кращих результатів. Як і очікувалось, найкращим розкладом зміни темпу навчання виявився косинусний з розігрівом та оновленнями. Цей розклад має мало параметрів, до того ж зручний у використанні. Загалом, використовуючи різні підходи, отримали точність 93,5 % на досить складному наборі даних, що на 22 % краще за базовий експеримент. У наступних дослідженнях планується розглянути покращення не лише моделі розпізнавання облич, а й детекції. Від якості детекції обличчя безпосередньо залежить точність розпізнавання.

Research on Face Recognition Based on Improved Convolutional Neural Network Using Raspberry Pi

In this paper, a face recognition system has been designed using Raspberry Pi based on convolution neural network technology. Convolution neural network that using OpenCV (open source computer vision library) to realize real-time face recognition and feature extraction. The difficult and important problem in deep learning network is that it could be over fitting very easily. In order to alleviate the over fitting phenomenon and improve the accuracy of face recognition, the dropout technology of convolutional neural network had been employed to improve the recognition accuracy and ensure the robustness of face recognition. The results of experiments show that our methods has the advantages in the computer time and recognition rate for the low construction cost, simple structure and strong portability. The accuracy of face detection for static and dynamic images using Raspberry Pi is above 85% in our testament.

Multimodal emotional analysis through hierarchical video summarization and face tracking

The era of video data has fascinated users into creating, processing, and manipulating videos for various applications. Voluminous video data requires higher computation power and processing time. In this work, a model is developed that can precisely acquire keyframes through hierarchical summarization and use the keyframes to detect faces and assess the emotional intent of the user. The key-frames are used to detect faces using recursive Viola-Jones algorithm and an emotional analysis for the faces extracted is conducted using an underlying architecture developed based on Deep Neural Networks (DNN). This work has significantly contributed in improving the accuracy of face detection and emotional analysis in non-redundant frames. The number of frames selected after summarization was less than 30% using the local minima extraction. The recursive routine introduced for face detection reduced false positives in all the video frames to lesser than 2%. The accuracy of emotional prediction on the faces acquired through the summarized frames, on Indian faces achieved a 90%. The computational requirement scaled down to 40% due to the hierarchical summarization that removed redundant frames and recursive face detection removed false localization of faces. The proposed model intends to emphasize the importance of keyframe detection and use them for facial emotional recognition.

Research on fatigue detection based on visual features

The high incidence of traffic accidents brings immeasurable losses to life. In order to avoid such crises, researchers and automakers have used many methods to solve this problem. Among them, technology based on visual features is widely used in driver fatigue detection. As fatigue detection plays a vital role in the driving process, the high accuracy of fatigue monitoring is very important. This paper focuses on the method based on convolutional neural network to detect driver fatigue. First, in the face detection part, the Single-Shot Multi-Box Detector algorithm is used to improve the speed and accuracy of face detection to extract the eye and mouth regions; second, the VGG16 network is used to learn fatigue features, which is performed on the NTHU-Drowsy Driver Detection (NTHU-DDD) data set and the other two modified data sets Training test. The main result of this work is that the accuracy of fatigue monitoring is higher than other methods including the original method, with an accuracy rate of over 90%. And it has better generalization ability than the multi-physical feature fusion detection method. At the same time, we propose the fatigue detection method based on convolutional neural network to improve the advanced driver assistance system (ADAS) to make it more robust and reliable decision making.

Research on Multi-Angle Face Detection Method Based on Improved YOLOV2 Algorithm

Face detection based on the YOLOV2 algorithm can achieve a higher level of accuracy and show a faster detection speed. Therefore, the YOLOV2 algorithm appears more frequently in the field of target real-time detection. The traditional YOLOV2 algorithm uses the Darknet-19 network. Based on this, this paper builds a new network by merging the residual network (ResNet). In addition, in order to further make up for the defects exposed by the YOLOV2 algorithm in face detection, we change the weight of the loss function and implement an image enhancement scheme on the training set. The YOLOV2 algorithm and the improved YOLOV2 algorithm are compared and tested on the Wider Face dataset, and then we use face data sets from different angles to test the performance of the algorithm. According to the analysis of the experimental results, the improved YOLOV2 algorithm can show higher accuracy and robustness in multi-angle face detection.

Face Detection Algorithm Based on Deep Residual Network

With the rapid development of artificial intelligence, face detection technology is widely used in our daily lives, such as mobile payment, video conferencing and personal identification. However, face the scenario while the face been blocking or crowding, the face detection accuracy would be greatly reduced. Therefore, in this paper, a high-precision face detection algorithm based on deep residual network has been proposed to solve this issue. Firstly, adding neural framework branches based on the Resnet-50 to improve the detection accuracy. Then import the soft-NMS method to enhance the robustness of algorithm. Experimental test on the public data set FDDB, the results indicate that the accuracy of this algorithm can reach 94.2% with good robustness, both accuracy and speed are better than the previous algorithm.

Face recognition in four types of colour space: a performance analysis

Face detection is one of the problems in computer vision with a large number of researches works. This research work aims to analyses the effect of performing various color spaces in face detection. 1100 images provided by SFA dataset was used in this study. The proposed method was initiated by pre-processing step. Then, segmentation step by using thresholding and morphological operation was conducted. For selecting face area, some features such as eccentricity, area and texture features were used. Finally, for validating the reliability of the proposed method, UTKface dataset was applied. This study obtained more than 90% accuracy of face detection in each color space. Moreover, the highest accuracy was obtained in YCbCr color space with 96.13%. This result was also strengthened with other experiment by using UTKface dataset. This experiment achieves accuracy of 80%. According to the performance analysis, the proposed method is reliable for recognizing face characteristic.

Custom Face Classification Model for Classroom Using Haar-Like and LBP Features with Their Performance Comparisons

Face detection by electronic systems has been leveraged by private and government establishments to enhance the effectiveness of a wide range of applications in our day to day activities, security, and businesses. Most face detection algorithms that can reduce the problems posed by constrained and unconstrained environmental conditions such as unbalanced illumination, weather condition, distance from the camera, and background variations, are highly computationally intensive. Therefore, they are primarily unemployable in real-time applications. This paper developed face detectors by utilizing selected Haar-like and local binary pattern features, based on their number of uses at each stage of training using MATLAB’s trainCascadeObjectDetector function. We used 2577 positive face samples and 37,206 negative samples to train Haar-like and LBP face detectors for a range of False Alarm Rate (FAR) values (i.e., 0.01, 0.05, and 0.1). However, the study shows that the Haar cascade face detector at a low stage (i.e., at six stages) for 0.1 FAR value is the most efficient when tested on a set of classroom images dataset with 100% True Positive Rate (TPR) face detection accuracy. Though, deep learning ResNet101 and ResNet50 outperformed the average performance of Haar cascade by 9.09% and 0.76% based on TPR, respectively. The simplicity and relatively low computational time used by our approach (i.e., 1.09 s) gives it an edge over deep learning (139.5 s), in online classroom applications. The TPR of the proposed algorithm is 92.71% when tested on images in the synthetic Labeled Faces in the Wild (LFW) dataset and 98.55% for images in MUCT face dataset “a”, resulting in a little improvement in average TPR over the conventional face identification system.

Real-time Face Mask-Wearing Detection and Temperature Measurement based on a Deep Learning Model

In the context of the COVID-19 outbreak in a global scenario, mandatory mask-wearing and temperature control can effectively prevent its spread and realize self-protection. Therefore, real-time face-mask wearing and temperature measurement technology is of greater importance against the background of infectious disease prevention and control. The present study adopted MobileNet as the backbone of the single-stage RetinaFace framework for real-time face detection and mask-wearing detection. Moreover, the focal loss function of α dynamic value was adopted to avoid the class imbalance problem and improve the classification accuracy in the training stage. Regarding face temperature measurement technology, non-contact and uncooled temperature-sensitive elements were used for temperature measurement, but it was easily affected by environmental variables. Therefore, an SVR model was employed for temperature calibration with the constant temperature blackbody as reference. The alignment errors for the accuracy of face detection, mask wearing detection and temperature correction were 89.58%, 97.84% and 4.85%, respectively. The parameter quantity of the face mask wearing detection model reached 0.42 M, while the computation quantity arrived at 2.039 GFLOPs. The detection model proposed in this study combines real-time mask-wearing detection with face temperature measurement, which can help to quickly measure the body temperature and detect whether one wears face masks properly in the context of COVID-19, so as to reduce the risk of epidemic spread.

Efficient Face Detection in the Fisheye Image Domain.

Significant progress has been made for face detection from normal images in recent years; however, accurate and fast face detection from fisheye images remains a challenging issue because of serious fisheye distortion in the peripheral region of the image. To improve face detection accuracy, we propose a light-weight location-aware network to distinguish the peripheral region from the central region in the feature learning stage. To match the face detector, the shape and scale of the anchor (bounding box) is made location dependent. The overall face detection system performs directly in the fisheye image domain without rectification and calibration and hence is agnostic of the fisheye projection parameters. Experiments on Wider-360 and real-world fisheye images using a single CPU core indeed show that our method is superior to the state-of-the-art real-time face detector RFB Net.

Automatic Identification of Individual Primates with Deep Learning Techniques.

SummaryThe difficulty of obtaining reliable individual identification of animals has limited researcher's ability to obtain quantitative data to address important ecological, behavioral, and conservation questions. Traditional marking methods placed animals at undue risk. Machine learning approaches for identifying species through analysis of animal images has been proved to be successful. But for many questions, there needs a tool to identify not only species but also individuals. Here, we introduce a system developed specifically for automated face detection and individual identification with deep learning methods using both videos and still-framed images that can be reliably used for multiple species. The system was trained and tested with a dataset containing 102,399 images of 1,040 individuals across 41 primate species whose individual identity was known and 6,562 images of 91 individuals across four carnivore species. For primates, the system correctly identified individuals 94.1% of the time and could process 31 facial images per second.

Prominent facial feature and hybrid learning method-based advanced face detector robust to head-up, head-down, and arbitrary rotation cases

Face detection is important for computer vision. Until now, the Viola and Jones algorithm is still the most popular face detector embedded in digital cameras. Although it has good performance in frontal face detection, it may not perform well in detecting rotated, head-up, and head-down faces. In this paper, we propose a more robust face detector based on prominent feature extraction. We find that, in addition to using eyes and mouth, which have been adopted by other algorithms, noses and ears are also important clues for face detection. Particularly, for a profile face, only a single eye can be detected. For head-up or head-down faces, eyes and mouth may not be well detected. However, in these cases, noses and ears can still be detected well and we can apply these features to improve the accuracy of rotated, head-up, and head-down face detection. To well detect these prominent features, in addition to the Viola–Jones detector, we also apply edge and color information and use the refocus algorithm based on the convolutional neural network. Simulations conducted on several popular multi-view face databases show that the proposed face detector can attain higher detection rates and is robust to rotated, head-up, and head-down cases.

Face Detection and Recognition Based on Visual Attention Mechanism Guidance Model in Unrestricted Posture

Performance of face detection and recognition is affected and damaged because occlusion often leads to missed detection. To reduce the recognition accuracy caused by facial occlusion and enhance the accuracy of face detection, a visual attention mechanism guidance model is proposed in this paper, which uses the visual attention mechanism to guide the model highlight the visible area of the occluded face; the face detection problem is simplified into the high-level semantic feature detection problem through the improved analytical network, and the location and scale of the face are predicted by the activation map to avoid additional parameter settings. A large number of simulation experiment results show that our proposed method is superior to other comparison algorithms for the accuracy of occlusion face detection and recognition on the face database. In addition, our proposed method achieves a better balance between detection accuracy and speed, which can be used in the field of security surveillance.

Real Time Automatic Attendance System for Face Recognition Using Face API and OpenCV

Traditionally, the attendance of students has been a major concern for the colleges and the faculty has to spend a lot of time and is a tedious job to mark attendance manually. Current biometric attendance system is not automatic that’s why wastes a lot of time, difficult to maintain and requires a queue for scanning fingerprints to mark their attendance. In Modern era everyone has Smartphone and connected via internet every time. In this paper attendance monitoring will be done through smart phone available with almost all faculty members. Some of popular object detection algorithms are back propagation neural network, region based convolution network (RCNN), faster RCNN, single shot detector. Our unified structure is based on YOLO V3 (You only look once) algorithm for face detection and Microsoft Azure using face API for face recognition (face database). The unique part is camera installed in classroom will take picture twice one at the start and one at the end to ensure students has attended complete class. YOLO V3 will first count the students in an image followed by identifying faces as known and unknown generating spreadsheets separately and an email is send at the end of month to students, parents and faculty. The designed system performs efficient in real time implementation for counting and detection. Our entire system has proven to gather high accuracy in face detection and performance.

Triple loss for hard face detection

Although face detection has been well addressed in the last decades, despite the achievements in recent years, effective detection of small, blurred and partially occluded faces in the wild remains a challenging task. Meanwhile, the trade-off between computational cost and accuracy is also an open research problem in this context. To tackle these challenges, in this paper, a novel context enhanced approach is proposed with structural optimization and loss function optimization. For loss function optimization, we introduce a hierarchical loss, referring to ``triple loss'' in this paper, to optimize the feature pyramid network (FPN) (Lin et al., 2017) based face detector. Additional layers are only applied during the training process. As a result, the computational cost is the same as FPN during inference. For structural optimization, we propose a context sensitive structure to increase the capacity of the prediction network to improve the accuracy of the output. In details, a three-branch inception subnet (Szegedy et al., 2015) based feature fusion module is employed to refine the original FPN without increasing the computational cost significantly, further improving low-level semantic information, which is originally extracted from a single convolutional layer in the backward pathway of FPN. The proposed approach is evaluated on two publicly available face detection benchmarks, FDDB and WIDER FACE. By using a VGG-16 based detector, experimental results indicate that the proposed method achieves a good balance between the accuracy and computational cost of face detection.

Improving the accuracy of old and young face detection in the template matching method with Fuzzy Associative Memory (FAM)

Every human has a face pattern and certain characteristics even though identical twins, but the human face pattern still has its own distinctiveness as well as old face patterns and young face patterns even though the human face pattern is very diverse but for young and old face patterns will be a difference between one face and the other face. Face detection (face detection) is one of the initial stages that very important in face recognition that is used in biometric identification. Face detection can also be used to search or index face data from images or videos that contain faces of various sizes, positions, and backgrounds. Face detection (face detection) automatically with the help of a computer is a problem that is not easy because the human face has a high level of variability both intra-personal and extra-personal variability. This study shows that systems with template matching methods combined with FAM can successfully detect differences in human faces, 80% accuracy, 10% better by using ordinary template matching.

Deep Neural Network for Better Face Processing

This Paper represents the face detection using advanced method deep neural network which uses deep learning frame work. The old models used to detect the faces were like Haar-cascade method which detect the faces with good approaches but there is some uncertainty in the accuracy of the old models, so in this system we will use the latest deep neural network model which is embedded with latest open cv and by using the deep learning model frame work which is weighted with some other files. By using this model, we can achieve the better accuracy in face detection which can be used for further purposes like auto focus in cameras, counting number of people etc. This model detects the faces accurately and paves the way for better recognition systems which can be used in many face biometric applications. For this purpose, low-cost computer board Raspberry Pi and Camera Sensor will be used.

Design and Implementation of Facial Recognition System for Visually Impaired using Image Processing

One of the most difficult tasks faced by the visually impaired students is identification of people. The rise in the field of image processing and the development of algorithms such as the face detection algorithm, face recognition algorithm gives motivation to develop devices that can assist the visually impaired. In this research, we represent the design and implementation of a facial recognition system for the visually impaired by using image processing. The device developed consists of a programmed raspberry pi hardware. The data is fed into the device in the form of images. The images are pre-processed and then the input image captured is processed inside the raspberry pi module using KNN algorithm, The face is recognized and the name is fed into text to speech conversion module. The visually impaired student will easily recognize the person before him using the device. Experiment results show high face detection accuracy and promising face recognition accuracy in suitable conditions. The device is built in such a way to improve cognition, interaction and communication of visually impaired students in schools and colleges. This system eliminates the need of a bulk computer since it employs a handy device with high processing power and reduced costs.

Application of face image detection based on deep learning in privacy security of intelligent cloud platform

The application of deep learning-based face detection in the privacy and security of intelligent cloud platforms is studied, in order to resolve the risk of private photo leakage on such platforms. Five key feature points of human faces (two eyes, two noses and two corners of mouth) are sought using the multitask cascaded convolution network (MTCNN). The algorithm utilizes the intrinsic links among Proposal Network (P-Net), Refine Network (R-Net) and Output Network (O-Net) to improve their face detection performances substantially. Since many existing data encryption methods, such as DES, RSA and AES, are applicable only to test data instead of digital images, the encryption of eigenvalues is achieved with a combination of chaotic logic diagrams and RC4 stream ciphers. Meanwhile, the MTCNN-generated face coordinates and user passwords are hash-converted and double-encrypted using hash table. The results show that the face detection accuracy of MTCNN reaches 95.04%, and that the image encryption method is suitable for network transmission. The chaotic logic diagrams increase the security of S-box initialization in the RC4 algorithm. The hash structure accelerates the file reading, whereas the hash conversion improves the security of critical data. In conclusion, the proposed encryption scheme is computationally fast and highly secure.

Performance Evaluation of MadBoost on Face Detection

Adaboost Viola-Jones method is indeed a profound discovery in detecting face images mainly because it is fast, light and one of the easiest methods of detecting face images among other techniques of face detection. Viola Jones uses Haar wavelet filter to detect face images and it produces almost 80%accuracy of face detection. This paper discusses proposed methodology and algorithms that involved larger library of filters used to create more discrimination features among the images by processing the proposed 15 Haar rectangular features (an extension from 4 Haar wavelet filters of Viola Jones) and used them in multiple adaptive ensemble process of detecting face image. After facial detection, the process continues with normalization processes by applying feature extraction such as PCA combined with LDA or LPP to extract our week learners’ wavelet for more classification features. Upon the process of feature extraction proposed feature selection to index these extracted data. These extracted vectors are used for training and creating MADBoost (Multiple Adaptive Diversified Boost)(an improvement of Adaboost, which uses multiple feature extraction methods combined with multiple classifiers) is able to capture, recognize and distinguish face image (s) faster. MADBoost applies the ensemble approach with better weights for classification to produce better face recognition results. Three experiments have been conducted to investigate the performance of the proposed MADBoost with three other classifiers, Neural Network (NN), Support Vector Machines (SVM) and Adaboost classifiers using Principal Component Analysis (PCA) as the feature extraction method. These experiments were tested against obstacles of POIES (Pose, Obstruction, Illumination, Expression, Sizes). Based on the results obtained, Madboost is found to be able to improve the recognition performance in matching failures, incorrect matching, matching success percentages and acceptable time taken to perform the classification task.