Face Recognition Task Research Articles

Hybrid Spatial-Channel Attention Mechanism for Cross-Age Face Recognition

Face recognition techniques have been widely employed in real-world biomimetics applications. However, traditional approaches have limitations in recognizing faces correctly with large age differences because of significant changes over age in the same person, leading to unsatisfactory recognition performance. To address this, previous studies propose to decompose and identify age and identity features independently in facial images across diverse age groups when optimizing the discriminative model so as to improve the age-invariant face recognition accuracy. Nevertheless, the interrelationships between these features make it difficult for the decomposition to disentangle them properly, thus compromising the recognition accuracy due to the interactive impacts on both features. To this end, this paper proposes a novel deep framework that incorporates a novel Hybrid Spatial-Channel Attention Module to facilitate the cross-age face recognition task. Particularly, the proposed module enables better decomposition of the facial features in both spatial and channel dimensions with attention mechanisms simultaneously while mitigating the impact of age variation on the recognition performance. Beyond this, diverse pooling strategies are also combined when applying those spatial and channel attention mechanisms, which allows the module to generate discriminative face representations while preserving complete information within the original features, further yielding sounder recognition accuracy. The proposed model is extensively validated through experiments on public face datasets such as CACD-VS, AgeDB-30, and FGNET, where the results show significant performance improvements compared to competitive baselines.

Differentiable Auxiliary Learning for Sketch Re-Identification

Sketch re-identification (Re-ID) seeks to match pedestrians' photos from surveillance videos with corresponding sketches. However, we observe that existing works still have two critical limitations: (i) cross- and intra-modality discrepancies hinder the extraction of modality-shared features, (ii) standard triplet loss fails to constrain latent feature distribution in each modality with inadequate samples. To overcome the above issues, we propose a differentiable auxiliary learning network (DALNet) to explore a robust auxiliary modality for Sketch Re-ID. Specifically, for (i) we construct an auxiliary modality by using a dynamic auxiliary generator (DAG) to bridge the gap between sketch and photo modalities. The auxiliary modality highlights the described person in photos to mitigate background clutter and learns sketch style through style refinement. Moreover, a modality interactive attention module (MIA) is presented to align the features and learn the invariant patterns of two modalities by auxiliary modality. To address (ii), we propose a multi-modality collaborative learning scheme (MMCL) to align the latent distribution of three modalities. An intra-modality circle loss in MMCL brings learned global and modality-shared features of the same identity closer in the case of insufficient samples within each modality. Extensive experiments verify the superior performance of our DALNet over the state-of-the-art methods for Sketch Re-ID, and the generalization in sketch-based image retrieval and sketch-photo face recognition tasks.

X2-Softmax: Margin adaptive loss function for face recognition

Learning the discriminative features of different faces is an important task in face recognition. By extracting face features with neural networks, it becomes easy to measure the similarity of different face images, which makes face recognition possible. To enhance a neural network’s face feature separability, incorporating an angular margin during training is common practice. The state-of-the-art loss functions CosFace and ArcFace apply fixed margins between the weights of classes to enhance the inter-class separation of face features. Since the distribution of samples in the training set is uneven, the similarities between different identities are unequal. Therefore, using an inappropriately fixed angular margin may lead to problems such as that the model has difficulty converging or that the face features are not sufficiently discriminative. It is more intuitive to use adaptive angular margins are angular adaptive, which can increase as the angles between classes increase. In this paper, we propose a new angular margin loss named X2-Softmax. X2-Softmax loss has adaptive angular margins, which increase as the angle between different classes increases. The angular adaptive margin ensures model flexibility and effectively improves the effect of face recognition. We trained a neural network with X2-Softmax loss on the MS1Mv3 dataset and tested it on several evaluation benchmarks to demonstrate the effectiveness and superiority of our loss function.

A study of Citespace-based deep learning applied to face recognition

Face recognition technology has made significant progress in recent years, driven by deep learning and other technologies, and is widely used in public safety, financial payment, intelligent access control, and other fields. Deep learning can effectively extract discriminative features from face images by constructing neural network structures and automatically learning feature mapping relationships, to improve the recognition accuracy. Deep learning shows excellent robustness when dealing with complex scenarios, and scientific metrology and visual analysis tools such as Citespace play an important role in analyzing the current status and development trend of applied research in the field of face recognition. Deep learning methods such as data enhancement techniques and generative adversarial networks have shown strong performance in face recognition tasks. In the future, the further integration and development of deep learning and face recognition technology will promote technological innovation and application expansion. Face recognition technology has important application potential in the digital society and will have wider application prospects in the future.

Spatial representation of multidimensional information in emotional faces revealed by fMRI

Face perception is a complex process that involves highly specialized procedures and mechanisms. Investigating into face perception can help us better understand how the brain processes fine-grained, multidimensional information. This research aimed to delve deeply into how different dimensions of facial information are represented in specific brain regions or through inter-regional connections via an implicit face recognition task. To capture the representation of various facial information in the brain, we employed support vector machine decoding, functional connectivity, and model-based representational similarity analysis on fMRI data, resulting in the identification of three crucial findings. Firstly, despite the implicit nature of the task, emotions were still represented in the brain, contrasting with all other facial information. Secondly, the connection between the medial amygdala and the parahippocampal gyrus was found to be essential for the representation of facial emotion in implicit tasks. Thirdly, in implicit tasks, arousal representation occurred in the parahippocampal gyrus, while valence depended on the connection between the primary visual cortex and the parahippocampal gyrus. In conclusion, these findings dissociate the neural mechanisms of emotional valence and arousal, revealing the precise spatial patterns of multidimensional information processing in faces.

A novel generative adversarial network‐based super‐resolution approach for face recognition

AbstractFace recognition is an essential feature required for a range of computer vision applications such as security, attendance systems, emotion detection, airport check‐in, and many others. The super‐resolution of subject images is an important and challenging element in numerous scenarios. At times the images are low resolution and need to be processed through super‐resolution techniques to gain more accurate results. For the problem of image super‐resolution, deep learning‐based face recognition systems have been explored in recent years; however, low‐resolution face recognition remains an arduous task. Generative adversarial network (GAN) based models are a promising approach to address this challenge. However, conventional GAN‐based models may generate images that differ significantly from an original high‐resolution image in the test set to the point that the identity of the target face may be changed. To address this shortcoming, we propose a novel U‐Net style generator architecture, where skip‐connections between the encoder and decoder layer can help in preserving the facial characteristics of the input image in the generated image, thus curbing the generator's ability to generate an entirely new image and training it to generate an image more similar in characteristics to the original image. In addition to statistical metrics like structural similarity index measure and Fréchet inception distance, we compute the pixel‐wise distance between the original and model‐generated images to ascertain that our model generates as close to the original images as possible. While we train the model for 4× super‐resolution (64 × 64 images to 256 × 256), our architecture can also be trained for an arbitrary resizing scale. Finally, the number of faces detected over high‐resolution images generated by our model is shown to be higher than state‐of‐the‐art high‐resolution image creation models for face recognition tasks.

Disrupted properties of functional brain networks in major depressive disorder during emotional face recognition: an EEG study via graph theory analysis.

Previous neuroimaging studies have revealed abnormal brain networks in patients with major depressive disorder (MDD) in emotional processing. While any cognitive task consists of a series of stages, little is yet known about the topology of functional brain networks in MDD for these stages during emotional face recognition. To address this problem, electroencephalography (EEG)-based functional brain networks of MDD patients at different stages of facial information processing were investigated in this study. First, EEG signals were collected from 16 patients with MDD and 18 age-, gender-, and education-matched normal subjects when performing an emotional face recognition task. Second, the global field power (GFP) method was employed to divide group-averaged event-related potentials into different stages. Third, using the phase transfer entropy (PTE) approach, the brain networks of MDD patients and normal individuals were constructed for each stage in negative and positive face processing, respectively. Finally, we compared the topological properties of brain networks of each stage between the two groups using graph theory approaches. The results showed that the analyzed three stages of emotional face processing corresponded to specific neurophysiological phases, namely, visual perception, face recognition, and emotional decision-making. It was also demonstrated that depressed patients showed abnormally decreased characteristic path length at the visual perception stage of negative face recognition and normalized characteristic path length in the stage of emotional decision-making during positive face processing compared to healthy subjects. Furthermore, while both the MDD and normal groups' brain networks were found to exhibit small-world network characteristics, the brain network of patients with depression tended to be randomized. Moreover, for patients with MDD, the centro-parietal region may lose its status as a hub in the process of facial expression identification. Together, our findings suggested that altered emotional function in MDD patients might be associated with disruptions in the topological organization of functional brain networks during emotional face recognition, which further deepened our understanding of the emotion processing dysfunction underlying MDD.

Automatic detection of nostril and key markers in images

In the medical field, one of complex challenges in surgery involving the skin, cartilage, mucosa, and the skeletal platform is the cleft nasal deformity. The causes of deformity are nasal formed from improper fusion of the medial and lateral nasal prominences with the maxillary prominence during embryologic development. Primary rhinoplasty or first time of cleft nose repair is the nasal surgery for repairing the cleft nasal deformity. In post cheiloplasty, the primary rhinoplasty patients must use nasal splint for supporting surgical wound. However, regular nasal splints are expensive and only come with default nasal size. This limits nasal splint from supporting specific patient cases such as children. The goal of this study is to implement a program that can automatically detect the nostril and green marker in patient images for creating custom-made 3D nasal splints. The proposed method is created utilizing the CNN model. YOLO architecture is utilized because it is the one of several CNN architectures that works well with the face recognition task. The YOLOV5, YOLOV8, and YOLO-NAS are attempted and compared in the training phase. The model with a highest performance is selected, and fine-tuned for adjusting to be compatible with a patient dataset. The fine-tuned YOLOV8 reaches a mAP with 99.5%. The predicted images from fine-tuned model are used to perform body part segmentation like Otsu's thresholding and discover contour to locate essential features like the green marker and nostril in ellipse and bounding boxes. The nostril distance is calculated using the bounding box and ellipse. The columella distance or nostril gap is measured from the distance between nostrils in YOLO predicted label. Both distances are converted into centimeter scale and evaluated with ground truth value in volunteer and patient cases from nasal expert for inspecting efficiency between bounding box, ellipse, and size of marker. The result proves that the proposed model can detect crucial features with mean absolute error 0.102. Furthermore, the proposed model indicates that marker size has no effect on detection and distance, according to the findings of the experiments.

Feasibility study of functional near-infrared spectroscopy in the ventral visual pathway for real-life applications.

fNIRS-based neuroenhancement depends on the feasible detection of hemodynamic responses in target brain regions. Using the lateral occipital complex (LOC) and the fusiform face area (FFA) in the ventral visual pathway as neurofeedback targets boosts performance in visual recognition. However, the feasibility of utilizing fNIRS to detect LOC and FFA activity in adults remains to be validated as the depth of these regions may exceed the detection limit of fNIRS. This study aims to investigate the feasibility of using fNIRS to measure hemodynamic responses in the ventral visual pathway, specifically in the LOC and FFA, in adults. We recorded the hemodynamic activities of the LOC and FFA regions in 35 subjects using a portable eight-channel fNIRS instrument. A standard one-back object and face recognition task was employed to elicit selective brain responses in the LOC and FFA regions. The placement of fNIRS optodes for LOC and FFA detection was guided by our group's transcranial brain atlas (TBA). Our findings revealed selective activation of the LOC target channel (CH2) in response to objects, whereas the FFA target channel (CH7) did not exhibit selective activation in response to faces. Our findings indicate that, although fNIRS detection has limitations in capturing FFA activity, the LOC region emerges as a viable target for fNIRS-based detection. Furthermore, our results advocate for the adoption of the TBA-based method for setting the LOC target channel, offering a promising solution for optrode placement. This feasibility study stands as the inaugural validation of fNIRS for detecting cortical activity in the ventral visual pathway, underscoring its ecological validity. We suggest that our findings establish a pivotal technical groundwork for prospective real-life applications of fNIRS-based research.

Improving Efficiency in Facial Recognition Tasks Through a Dataset Optimization Approach

Improving Efficiency in Facial Recognition Tasks Through a Dataset Optimization Approach

'Magic cosmetic fillers': Appearance-enhancement effects on self-face recognition.

People naturally exhibit a self-serving bias which can be observed in their tendency to judge their own physical attractiveness more favourably than that of others. Despite this positive self-perception, minimally invasive cosmetic injectable procedures for facial rejuvenation and enhancement are becoming increasingly common. It remains unclear, however, whether recognizing an altered version of one's own face, enhanced cosmetically, correlates with a positive view of cosmetic surgery and excessive preoccupations about physical characteristics perceived as defects (body dysmorphic concerns). In this study, 30 healthy female participants, aged 18-24 years (Mage = 21.1 years, SD = 1.6), engaged in a face recognition task during which their faces were digitally morphed with that of gender-matched unfamiliar women who had undergone cosmetic enhancements, specifically lip and cheek fillers. The duration of exposure to these modified faces varied with short (500 msec) and long (2000 msec) viewing periods. Participants were asked to identify whether the digital morphs represented themselves or the other woman. Self-reports regarding acceptance of cosmetic surgery and dysmorphic concerns were collected. Participants PSE indicated a tendency towards self-bias under short presentation times, shifting towards the other as presentation times lengthened. Interestingly, this effect was associated with greater acceptance of cosmetic surgery and higher body dysmorphic concerns. This study underscores the importance of understanding how perceptions of others' physical appearances can influence self-recognition and attitudes towards cosmetic surgery, which may have both positive and potentially harmful implications.

Accelerating automatic model finding with layer replications case study of MobileNetV2.

In this paper, we propose a method to reduce the model architecture searching time. We consider MobileNetV2 for 3D face recognition tasks as a case study and introducing the layer replication to enhance accuracy. For a given network, various layers can be replicated, and effective replication can yield better accuracy. Our proposed algorithm identifies the optimal layer replication configuration for the model. We considered two acceleration methods: distributed data-parallel training and concurrent model training. Our experiments demonstrate the effectiveness of the automatic model finding process for layer replication, using both distributed data-parallel and concurrent training under different conditions. The accuracy of our model improved by up to 6% compared to the previous work on 3D MobileNetV2, and by 8% compared to the vanilla MobileNetV2. Training models with distributed data-parallel across four GPUs reduced model training time by up to 75% compared to traditional training on a single GPU. Additionally, the automatic model finding process with concurrent training was 1,932 minutes faster than the distributed training approach in finding an optimal solution.

MASKED AND UNMASKED FACE RECOGNITION ON UNCONSTRAINED FACIAL IMAGES USING HAND-CRAFTED METHODS

In this study, the face recognition task is applied on masked and unmasked faces using hand-crafted methods. Due to COVID-19 and masks, facial identification from unconstrained images became a hot topic. To avoid COVID-19, most people use masks outside. In many cases, typical facial recognition technology is useless. The majority of contemporary advanced face recognition methods are based on deep learning, which primarily relies on a huge number of training examples, however, masked face recognition may be investigated using hand-crafted approaches at a lower computing cost than using deep learning systems. A low-cost system is intended to be constructed for recognizing masked faces and compares its performance to that of face recognition systems that do not use masks. The proposed method fuses hand-crafted methods using feature-level fusion strategy. This study compares the performance of masked and unmasked face recognition systems. The experiments are undertaken on two publicly accessible datasets for masked face recognition: Masked Labeled Faces in the Wild (MLFW) and Cross-Age Labeled Faces in the Wild (CALFW). The best accuracy is achieved as 94.8% on MLFW dataset. The rest of the results on different train and test sets from CALFW and MLFW datasets are encouraging compared to the state-of-the-art models.

Max-C and Min-D Projection Auto-Associative Fuzzy Morphological Memories: Theory and an Application for Face Recognition

Max-C and min-D projection auto-associative fuzzy morphological memories (max-C and min-D PAFMMs) are two-layer feedforward fuzzy morphological neural networks designed to store and retrieve finite fuzzy sets. This paper addresses the main features of these auto-associative memories: unlimited absolute storage capacity, fast retrieval of stored items, few spurious memories, and excellent tolerance to either dilative or erosive noise. Particular attention is given to the so-called Zadeh’ PAFMM, which exhibits the most significant noise tolerance among the max-C and min-D PAFMMs besides performing no floating-point arithmetic operations. Computational experiments reveal that Zadeh’s max-C PFAMM, combined with a noise masking strategy, yields a fast and robust classifier with a strong potential for face recognition tasks.

Degradation model and attention guided distillation approach for low resolution face recognition

Deep convolution neural networks (CNN) have shown their efficacy in face recognition tasks due to their ability to extract highly discriminant face representations from face images. On high-resolution benchmark datasets, outstanding identification and verification results have been achieved. However, the performance of these algorithms is significantly degraded when tested on low-resolution (LR) images such as those captured from surveillance cameras. A straightforward solution to this problem is to use both high-resolution (HR) images and corresponding down-sampled LR images during training. Although this strategy improves the performance for LR images, it has some limitations. First, there is a significant difference between down-sampled LR images and LR images from surveillance cameras, leading to performance saturation at an earlier stage. Another limitation is the deterioration in the performance of HR images. In this work, solutions to both these limitations are proposed. A degradation model is proposed that synthesizes LR images from corresponding HR, emulating the real-world degradation effects in synthetic data, thus enabling the face recognition system to tolerate various blurry and noisy effects. To address the deterioration in the performance of HR images, an attention-guided distillation is proposed, which utilizes attention maps from convolutional layers in combination with deep features to transfer informative HR features from teacher to student network. The attention maps from the teacher network guide the student network to a better optimum and produce resolution robust face representations. The results of the proposed approach on the popular LR datasets like SCface, Coxface, and PaSC show that it outperforms the recent state-of-the-art (SOTA) techniques by a significant margin demonstrating its effectiveness for different cross-resolution scenarios.

Application of Multi-sensory interaction design Based on Machine Learning in Virtual Reality

Virtual reality (VR) technology is well-known to the public for its unparalleled sense of environmental immersion, which is accompanied by a large amount of human-computer interaction between users and objects in the virtual environment. Users transmit and obtain information through their senses, so the processing mechanism of information in various senses under the background of VR technology has become the foundation of research. In the process of natural information dissemination, users’ senses are synergistically involved, so whether there is a correlation between various senses that affects users’ final perception has become an important theoretical support for studying multi-sensory interaction. Facial expression recognition is an important component of multi-sensory interaction technology. Therefore, this article mainly uses machine learning algorithms to achieve real-time detection of facial targets in images or videos, as well as classification and recognition of facial expressions. To meet the task requirements of the expression recognition module, this article selects the DeepID network model based on deep learning in the field of facial recognition as the basic model for the expression recognition function. This study utilizes the excellent feature extraction ability of the original network for facial images and introduces residual modules and attention mechanisms, making the modified network model more suitable for facial expression multi classification recognition tasks. The recognition accuracy of the facial expression classification algorithm proposed by this work reaches 97.2%.

Cyclic style generative adversarial network for near infrared and visible light face recognition

Near Infrared and Visible Light (NIR-VIS) face recognition attracts attention from researchers because of its potential for safety, illumination invariance, and stability. Nevertheless, the difference between NIR and VIS domains, the domain gap, remains a huge problem for matching NIR and VIS images. Specifically, for the same identity, the appearance in the NIR domain is different from the VIS domain. Thus, traditional face recognition methods cannot be effective. To address this problem, this paper proposes a novel model, called Cyclic-Style GAN (CS-GAN). First, the pre-trained Style-GAN 3 network is embedded into the Cycle-GAN structure for NIR-VIS cross-domain learning. Second, there is a cyclic subspace learning method consisting of latent loss and style loss, through which both style (domain feature) and facial characteristic features are learned to improve the quality of synthesized images. The model synthesizes realistic VIS images from NIR images and does the face recognition task in the VIS domain. The proposed method achieves 99.6% Rank-1 accuracy on the CASIA NIR-VIS 2.0 database which is a state-of-the-art result. The visualization results show that the proposed model synthesizes VIS images with a clear texture of faces and in close-to-reality color.

Emotion regulation effects of Ayahuasca in experienced subjects during implicit aversive stimulation: An fMRI study

Ethnopharmacological relevanceAyahuasca is a beverage used in Amazonian traditional medicine and it has been part of the human experience for millennia as well as other different psychoactive plants. Although Ayahuasca has been proposed as potentially therapeutic as an anxiolytic and antidepressant, whilst no studies have been carried out so far investigating their direct effect on brain emotional processing. Aim of the studyThis study aimed to measure the emotional acute effect of Ayahuasca on brain response to implicit aversive stimulation using a face recognition task in functional magnetic resonance imaging (fMRI). Materials and methodsNineteen male experienced Ayahuasca users participated in this study in two fMRI sessions before and after 50 min of the Ayahuasca ingestion. Subjects were presented with pictures of neutral (A) and aversive (B) (fearful or disgusted) faces from the Pictures of Facial Affect Series. Subjects were instructed to identify the gender of the faces (gender discrimination task) while the emotional content was implicit. Subjective mood states were also evaluated before Ayahuasca intake and after the second fMRI session, using a visual analogue mood scale (VAMS). ResultsDuring the aversive stimuli, the activity in the bilateral amygdala was attenuated by Ayahuasca (qFDR<0.05). Furthermore, in an exploratory analysis of the effects after intake, Ayahuasca enhances the activation in the insular cortex bilaterally, as well as in the right dorsolateral prefrontal cortex (qFDR<0.05). In the psychometric VAMS scale, subjects reported attenuation of both anxiety and mental sedation (p < 0.01) during acute effects. ConclusionsTogether, all reported results including neuroimaging, behavioral data and psychometric self-report suggest that Ayahuasca can promote an emotion regulation mechanism in response to aversive stimuli with corresponding improved cognition including reduced anxiety and mental sedation.

Self-supervised Learning for Expression Recognition on Small-scale Data Set

In this paper, we propose a facial micro-expression recognition method utilizing self-supervised learning and Vision Transformer. We employ a contrast learning approach for self-supervision and extract image features using an attention mechanism. Various data augmentation techniques were utilized, and we specifically designed an enhancement method for facial recognition. By combining the strengths of the Vision Transformer and CNN models for feature extraction, our approach achieves improved recognition accuracy, even with limited labeled data. Experimental evaluation shows that our proposed method has good results in facial micro-expression recognition tasks.

Can machine recognize a long-missed old friend? A test to the FaceNet face recognition algorithm

Facial recognition receives heated discussion in recent years. Given the increasing demand for portable and easy-to-use face recognition programs, several coding libraries have been developed that includes the algorithm for identifying and classifying faces. One example of such libraries, the “face_recognition” library developed by Adam Geitgey allows fast execution for face recognition tasks and requires only 1 photo as training input. It adopts the Convolutional Neural Network (CNN) from OpenFace, which is built based on the FaceNet algorithm. Although the study claims that the algorithm has a high base on tests from common face dataset, it is not clear whether the algorithm can identify people consistently when they are in their different ages. In this paper, a face dataset is constructed by collecting publicly available photos from celebrities that can be traced to a certain date of taken. These photos are grouped by 5 pieces per person with each photo taken roughly 4 years apart from the previous one. Then, experiments are conducted to test the effectiveness of face_recognition library on identifying the person from these groups of photos. The results report an over 75% success rate for the algorithm on image identification tasks, indicating a good ability for the face encoding to resist the gap of time.