Development Of Face Recognition Research Articles

Research on occlusion face recognition based on deep networks

Based on the spread of the new crown epidemic, the use of masks has been popularized, so it has a significant impact on the development of face recognition under the cover. The study of how to improve the performance of face recognition under occlusion conditions is also an important topic in the field of face recognition in the future. At the same time, the neural network model is one of the most important models in deep learning, in the field of image classification, face recognition based on deep network has also been proved to be an efficient feature extraction method, this paper divides the face recognition method based on occlusion into two categories: local feature class based on non-occlusion area and feature class based on recognition occlusion area; The basic processes of these two types of methods are summarized, and the specific cases of these two types of occlusion face recognition methods are analyzed. Further summarize the advantages and disadvantages of each and elaborate them. At the end of the article, the shortcomings and future development trends of the current shielding face recognition research are summarized.

Localizing Face Recognition with Haar-Cascade Classifier and LBPH using Python

This paper experimented with the combination of the Haar-Cascade Classifier and Local Binary Pattern Histogram in creating a Face Recognition System using Python programming language. Images were gathered using Python script from the 10persons selected while facing the camera. Each individual had 1000 photos that were used to train a model. Another Python script was created for training, identification and recognition. The overall efficiency rating of the model was 84% which denotes a practical recommendation for utilization. The findings of this study contribute as a reference for further development of face recognition with other image classification algorithms

Exploring The Efficiency of Resnet and Densenet in Gender Recognition

With the development of face recognition and image classification. There are many scenarios where gender classification techniques need to be applied. For example, the stores can recommend different products according to the gender of customers. There are several methods of gender classification, one of them is the convolutional neural network (CNN), which shows its promising potential in image classification. With other relevant variables and parameters being the same, this work tries to compare the Resnet with the Densenet to find a model that is suitable for specific gender classification datasets. The classification efficiency of the model is examined from the three directions of running prediction loss. This paper finds that in the case of small-scale data sets, the classification performanceof the Densenet model ourperforms that of the Resnet model, and the number of parameters used by the Densenet is significantly less than that of the Resnet. When a high-precision classification or recognition model is needed, Densenet is preferred.

GicoFace: A Deep Face Recognition Model Based on Global-Information Loss Function

As CNNs have a strong capacity to learn discriminative facial features, CNNs have greatly promoted the development of face recognition, where the loss function plays a key role in this process. Nonetheless, most of the existing loss functions do not simultaneously apply weight normalization, apply feature normalization and follow the two goals of enhancing the discriminative capacity (optimizing intra-class/inter-class variance). In addition, they are updated by only considering the feedback information of each mini-batch, but ignore the information from the entire training set. This paper presents a new loss function called Gico loss. The deep model trained with Gico loss in this paper is then called GicoFace. Gico loss satisfies the four aforementioned key points, and is calculated with the global information extracted from the entire training set. The experiments are carried out on five benchmark datasets including LFW, SLLFW, YTF, MegaFace and FaceScrub. Experimental results confirm the efficacy of the proposed method and show the state-of-the-art performance of the method.

Distorting Face Representations in Newborn Brains.

What role does experience play in the development of face recognition? A growing body of evidence indicates that newborn brains need slowly changing visual experiences to develop accurate visual recognition abilities. All of the work supporting this "slowness constraint" on visual development comes from studies testing basic-level object recognition. Here, we present the results of controlled-rearing experiments that provide evidence for a slowness constraint on the development of face recognition, a prototypical subordinate-level object recognition task. We found that (1) newborn chicks can rapidly develop view-invariant face recognition and (2) the development of this ability relies on experience with slowly moving faces. When chicks were reared with quickly moving faces, they built distorted face representations that largely lacked invariance to viewpoint changes, effectively "breaking" their face recognition abilities. These results provide causal evidence that slowly changing visual experiences play a critical role in the development of face recognition, akin to basic-level object recognition. Thus, face recognition is not a hardwired property of vision but is learned rapidly as the visual system adapts to the temporal structure of the animal's visual environment.

Lightened SphereFace for face recognition

Convolutional neural networks (CNN) have immensely promoted the development of face recognition (FR) technology. In order to achieve global accuracy, CNN models tend to be deeper or multiple local facial patch ensembles, leading to excessive amounts of calculation. We address these deep FR problems and propose a lightened deep learning framework under an open-set protocol to achieve a good classification effect and streamline the model itself. To this end, we improve the SphereFace that enables the deep network to learn angularly discriminative features more efficiently. First, global average pooling (GAP) is introduced to replace the original fully connected (FC) layer, which greatly reduces the storage of the model. Compared to the widely used FC layer, GAP can reduce the number of parameters and avoid overfitting. Then multilayer perceptron is added between convolution layers, which increases the ability to characterize features. These models are trained on the CASIA-WebFace dataset and evaluated on the Labeled Faces in the Wild and YTF datasets, which show the excellent performance of lightened SphereFace (L-SphereFace) in FR tasks. At the same time, computational cost is reduced in comparison with the released SphereFace model. The storage space of the model is also greatly compressed.

High Performance and Efficient Real-Time Face Detector on Central Processing Unit Based on Convolutional Neural Network

Face detection is crucial in the development of face recognition, expression, tracking, and classification. Conventional methods have accuracy constraints on several challenging conditions, including nonfrontal faces, occlusions, and complex backgrounds. However, the convolutional neural network (CNN) methods produce high performances despite a large amount of computation. Therefore, CNN requires expensive hardware and is not suitable for low-cost central processing units (CPUs). This article develops a light architecture for a CNN-based real-time face detector. The proposed architecture consists of two main modules, the backbone to extract distinctive facial features and multilevel detection to perform prediction at multiple scales. Furthermore, it utilizes several approaches to enhance the training result, including balancing loss and tweaks on the training configuration. The proposed detector has one stage and is trained using the input of images from WIDER FACE with challenges, which contains more challenging images than other datasets. As a result, the detector achieves state-of-the-art performance on several benchmark datasets compared with the other CPU-based models. Then, its efficiency is superior to that of competitors, as it runs at 53 frames per second on a CPU for video graphics array resolution images.

Multi-Pose Face Recognition Based on Deep Learning in Unconstrained Scene

At present, deep learning drives the rapid development of face recognition. However, in the unconstrained scenario, the change of facial posture has a great impact on face recognition. Moreover, the current model still has some shortcomings in accuracy and robustness. The existing research has formulated two methods to solve the above problems. One method is to model and train each pose separately. Then, a fusion decision will be made. The other method is to make “frontal” faces on the image or feature level and transform them into “frontal” face recognition. Based on the second idea, we propose a profile to the frontal revise mapping (PTFRM) module. This module realizes the revision of arbitrary poses on the feature level and transforms the multi-pose features into an approximate frontal representation to enhance the recognition ability of the existing recognition models. Finally, we evaluate the PTFRM on unconstrained face validation benchmark datasets such as Labeled Faces in the Wild (LFW), Celebrities in Frontal Profile (CFP), and IARPA Janus Benchmark A(IJB-A). Results show that the chosen method for this study achieves good performance.

Children are sensitive to mutual information in intermediate-complexity face and non-face features.

Understanding developmental changes in children's use of specific visual information for recognizing object categories is essential for understanding how experience shapes recognition. Research on the development of face recognition has focused on children's use of low-level information (e.g. orientation sub-bands), or high-level information. In face categorization tasks, adults also exhibit sensitivity to intermediate complexity features that are diagnostic of the presence of a face. Do children also use intermediate complexity features for categorizing faces and objects, and, if so, how does their sensitivity to such features change during childhood? Intermediate-complexity features bridge the gap between low- and high-level processing: they have computational benefits for object detection and segmentation, and are known to drive neural responses in the ventral visual system. Here, we have investigated the developmental trajectory of children's sensitivity to diagnostic category information in intermediate-complexity features. We presented children (5–10 years old) and adults with image fragments of faces (Experiment 1) and cars (Experiment 2) varying in their mutual information, which quantifies a fragment's diagnosticity of a specific category. Our goal was to determine whether children were sensitive to the amount of mutual information in these fragments, and if their information usage is different from adults. We found that despite better overall categorization performance in adults, all children were sensitive to fragment diagnosticity in both categories, suggesting that intermediate representations of appearance are established early in childhood. Moreover, children's usage of mutual information was not limited to face fragments, suggesting the extracting intermediate-complexity features is a process that is not specific only to faces. We discuss the implications of our findings for developmental theories of face and object recognition.

Review on Deep Learning-Based Face Analysis

This paper reviews the development of face recognition based on deep learning in the field of biometrics. Firstly, the basic application of face recognition and the definition of the deep learning model is explained. In addition, the research overview and application are summarized, such as face recognition method based on convolution neural network (CNN), deep nonlinear face shape extraction method, face-based robustness modeling based on deep learning, fully automatic face recognition in constrained environments, face recognition based on deep learning video monitoring, low resolution face recognition based on deep learning, and other deep learning of the face information recognition; analysis of the current face recognition technology in the deep learning applications in the problems and development trends. Finally, it is concluded that the deep learning can learn to get more useful data and can build a more accurate model. However, there are some shortcomings in deep learning, such as the length of the training model, the need for continuous iteration to model optimization, being difficult to guarantee the optimal global solution, which also needs to continue to explore in the future.

Development of Face Recognition on Raspberry Pi for Security Enhancement of Smart Home System

Development of Face Recognition on Raspberry Pi for Security Enhancement of Smart Home System

Development of Face Recognition on Raspberry Pi for Security Enhancement of Smart Home System

Nowadays, there is a growing interest in the smart home system using Internet of Things. One of the important aspect in the smart home system is the security capability which can simply lock and unlock the door or the gate. In this paper, we proposed a face recognition security system using Raspberry Pi which can be connected to the smart home system. Eigenface was used the feature extraction, while Principal Component Analysis (PCA) was used as the classifier. The output of face recognition algorithm is then connected to the relay circuit, in which it will lock or unlock the magnetic lock placed at the door. Results showed the effectiveness of our proposed system, in which we obtain around 90% face recognition accuracy. We also proposed a hierarchical image processing approach to reduce the training or testing time while improving the recognition accuracy.

Development of face recognition: Dynamic causal modelling of MEG data

Electrophysiological studies of adults indicate that brain activity is enhanced during viewing of repeated faces, at a latency of about 250 ms after the onset of the face (M250/N250). The present study aimed to determine if this effect was also present in preschool-aged children, whose brain activity was measured in a custom-sized pediatric MEG system. The results showed that, unlike adults, face repetition did not show any significant modulation of M250 amplitude in children; however children’s M250 latencies were significantly faster for repeated than non-repeated faces. Dynamic causal modelling (DCM) of the M250 in both age groups tested the effects of face repetition within the core face network including the occipital face area (OFA), the fusiform face area (FFA), and the superior temporal sulcus (STS). DCM revealed that repetition of identical faces altered both forward and backward connections in children and adults; however the modulations involved inputs to both FFA and OFA in adults but only to OFA in children. These findings suggest that the amplitude-insensitivity of the immature M250 may be due to a weaker connection between the FFA and lower visual areas.

Face-space: A unifying concept in face recognition research.

The concept of a multidimensional psychological space, in which faces can be represented according to their perceived properties, is fundamental to the modern theorist in face processing. Yet the idea was not clearly expressed until 1991. The background that led to the development of face-space is explained, and its continuing influence on theories of face processing is discussed. Research that has explored the properties of the face-space and sought to understand caricature, including facial adaptation paradigms, is reviewed. Face-space as a theoretical framework for understanding the effect of ethnicity and the development of face recognition is evaluated. Finally, two applications of face-space in the forensic setting are discussed. From initially being presented as a model to explain distinctiveness, inversion, and the effect of ethnicity, face-space has become a central pillar in many aspects of face processing. It is currently being developed to help us understand adaptation effects with faces. While being in principle a simple concept, face-space has shaped, and continues to shape, our understanding of face perception.

Face recognition based on subset selection via metric learning on manifold

With the development of face recognition using sparse representation based classification (SRC), many relevant methods have been proposed and investigated. However, when the dictionary is large and the representation is sparse, only a small proportion of the elements contributes to the l 1-minimization. Under this observation, several approaches have been developed to carry out an efficient element selection procedure before SRC. In this paper, we employ a metric learning approach which helps find the active elements correctly by taking into account the interclass/intraclass relationship and manifold structure of face images. After the metric has been learned, a neighborhood graph is constructed in the projected space. A fast marching algorithm is used to rapidly select the subset from the graph, and SRC is implemented for classification. Experimental results show that our method achieves promising performance and significant efficiency enhancement.

That’s my teacher! Children’s ability to recognize personally familiar and unfamiliar faces improves with age

Most previous research on the development of face recognition has focused on recognition of highly controlled images. One of the biggest challenges of face recognition is to identify an individual across images that capture natural variability in appearance. We created a child-friendly version of Jenkins, White, Van Montford, and Burton’s sorting task (Cognition, 2011, Vol. 121, pp. 313–323) to investigate children’s recognition of personally familiar and unfamiliar faces. Children between 4 and 12years of age were presented with a familiar/unfamiliar teacher’s house and a pile of face photographs (nine pictures each of the teacher and another identity). Each child was asked to put all the pictures of the teacher inside the house while keeping the other identity out. Children over 6years of age showed adult-like familiar face recognition. Unfamiliar face recognition improved across the entire age range, with considerable variability in children’s performance. These findings suggest that children’s ability to tolerate within-person variability improves with age and support a face-space framework in which faces are represented as regions, the size of which increases with age.

Typical and Atypical Development of Functional Connectivity in the Face Network.

Extensive studies have demonstrated that face recognition performance does not reach adult levels until adolescence. However, there is no consensus on whether such prolonged improvement stems from development of general cognitive factors or face-specific mechanisms. Here, we used behavioral experiments and functional magnetic resonance imaging (fMRI) to evaluate these two hypotheses. With a large cohort of children (n = 379), we found that the ability of face-specific recognition in humans increased with age throughout childhood and into late adolescence in both face memory and face perception. Neurally, to circumvent the potential problem of age differences in task performance, attention, or cognitive strategies in task-state fMRI studies, we measured the resting-state functional connectivity (RSFC) between the occipital face area (OFA) and fusiform face area (FFA) in human brain and found that the OFA-FFA RSFC increased until 11-13 years of age. Moreover, the OFA-FFA RSFC was selectively impaired in adults with developmental prosopagnosia (DP). In contrast, no age-related changes or differences between DP and normal adults were observed for RSFCs in the object system. Finally, the OFA-FFA RSFC matured earlier than face selectivity in either the OFA or FFA. These results suggest the critical role of the OFA-FFA RSFC in the development of face recognition. Together, our findings support the hypothesis that prolonged development of face recognition is face specific, not domain general.

System Development of Face Recognition on Android Platform and Optimized Information Provision Service with Server Network for Customs in Optician"s

System Development of Face Recognition on Android Platform and Optimized Information Provision Service with Server Network for Customs in Optician"s

The Other-Race Effect in Caucasian and Japanese-Descendant Children in Brazil: Evidence of Developmental Plasticity

The Other-Race Effect has been confirmed by several experimental studies, in which the individual has greater difficulty in recognizing the faces of races that are different from their own. Few studies have investigated this effect during the development of the face processing system. The aim of this study was to investigate the development of the Other-Race Effect in Caucasian and Japanese-descendants children born and living in Brazil. Seventy-four children, split into two age groups (5 - 7 and 9 - 11 years of age), were tested. Japanese-descendant children did not present the effect in favor of their own-race faces, whereas Caucasian children demonstrated the effect in both age groups. This indicates that the effect is present early in the development of face recognition and that contact with the faces of another race during childhood dissipates it. These findings suggest that experience with faces from the children’s visual context is crucial for shaping face processing.

Development of Face Recognition in 5- to 15-Year-Olds

This study focuses on the development of face recognition in typically developing preschool- and school-aged children (aged 5 to 15 years old, n = 611, 336 girls). Social predictors include sex differences and own-sex bias. At younger ages, the development of face recognition was rapid and became more gradual as the age increased up until the age of 11, after which point the influence of age was insignificant. This development could not be sufficiently explained by the improvement in visual attention or design memory tasks. Girls were slightly better than boys at recognizing faces in the youngest age group, but the effect of sex was minor. No significant own-sex bias was found for girls or boys.