Search

Facial expression recognition (FER) has advanced applications in various disciplines, including computer vision, Internet of Things, and artificial intelligence, supporting diverse domains such as medical escort services, learning analysis, fatigue detection, and human-computer interaction. The accuracy of these systems is of utmost concern and depends on effective feature selection, which directly impacts their ability to accurately detect facial expressions across various poses. This research proposes a new hybrid approach called QIFABC (Hybrid Quantum-Inspired Firefly and Artificial Bee Colony Algorithm), which combines the Quantum-Inspired Firefly Algorithm (QIFA) with the Artificial Bee Colony (ABC) method to enhance feature selection for a multi-pose facial expression recognition system. The proposed algorithm uses the attributes of both the QIFA and ABC algorithms to enhance search space exploration, thereby improving the robustness of features in FER. The firefly agents initially move toward the brightest firefly until identified, then search transition to the ABC algorithm, targeting positions with the highest nectar quality. In order to evaluate the efficacy of the proposed QIFABC algorithm, feature selection is also conducted using QIFA, FA, and ABC algorithms. The evaluated features are utilized for classifying face expressions by utilizing the deep neural network model, ResNet-50. The presented FER system has been tested using multi-pose facial expression benchmark datasets, including RaF (Radboud Faces) and KDEF (Karolinska Directed Emotional Faces). Experimental results show that the proposed QIFABC with ResNet50 method achieves an accuracy of 98.93%, 94.11%, and 91.79% for front, diagonal, and profile poses on the RaF dataset, respectively, and 98.47%, 93.88%, and 91.58% on the KDEF dataset., Competing Interests: Declarations. Competing interests: The authors declare no competing interests. Conflict of interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (© 2025. The Author(s).)

Background: Patients with bipolar disorders (BD) have significant impairments in Facial Expression Recognition (FER), an essential social skill for effective social interactions. While the Facial Emotions Recognition Test, 54 photographs by Gaudelus (TREF-54g) has been used in patients with schizophrenia, no study has evaluated FER using this test in BD patients. The TREF-54g meets the International Society for Bipolar Disorders (ISBD) criteria for assessing FER. The objective of this study was to compare FER in patients with euthymic BD versus healthy controls (HC) using the TREF-54g., Methods: This study included 60 patients with BD diagnosed according to DSM-5 criteria, 30 with bipolar I disorder (BDI) and 30 with bipolar II disorder (BDII) and 60 HC. All groups were matched for age, gender, and years of education. Overall TREF-54g impairment score was defined as a difference of two SD from the group of HC with mean overall accuracy score expressed as a percentage of correct responses., Results: Overall TREF-54g impairment scores, as well as sub-scores for anger and contempt recognition, were significantly higher in patients with BD versus those in HC. No significant differences in TREF-54g scores were shown between patients with BDI and those with BDII., Limitations: Clinical assessments of this study didn't include neurocognition assessments and classification of contempt emotion as a basic emotion remains controversial., Conclusions: We found significant impairments in FER in patients with euthymic BD, particularly for anger and contempt using the TREF-54g a test that meets ISBD recommendation criteria., Competing Interests: Declaration of competing interest Alain Dervaux has received honoraria for lectures and non-financial support from Camurus, Indivior, Janssen, Otsuka, and Lundbeck, outside the submitted work. Marie-Cecile Bralet has received honoraria for lectures and non-financial support from Boehringer, Otsuka, and Lundbeck, outside the submitted work. All other authors declare that they have no conflicts of interest. All other authors declare that they have no conflicts of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Facial expression recognition (FER) in the wild is a challenging pattern recognition task affected by the images' low quality and has attracted broad interest in computer vision. Existing FER methods failed to obtain sufficient accuracy to support the practical applications, especially in scenarios with low fault tolerance, which limits the adaptability of FER. Targeting exploring the possibility of further improving the accuracy of FER in the wild, this paper proposes a novel single model named R18+FAML and an ensemble model named R18+FAML-FGA-T2V, which applies intra-feature fusion within a single network, feature fusion among multiple networks, and the ensemble decision strategy. Based on the backbone of ResNet18 (R18), R18+FAML combines internal feature fusion and three attention blocks, as well as uses multiple loss functions (FAML) to improve the diversity of the feature extraction. To effectively integrate feature extractors from multiple networks, we propose feature fusion among networks based on the genetic algorithm (FGA). Comprehensively considering and utilizing more classification information, we propose an ensemble strategy, i.e., the improved top-two-voting (T2V) of multiple networks with the same structure. Combining the above strategies, R18+FAML-FGA-T2V can focus on the main expression-aware areas by integrating interest areas of multiple networks. From experiments on three challenging FER datasets in the wild including RAF-DB, AffectNet-8 and AffectNet-7, our single model R18+FAML and ensemble model R18+FAML-FGA-T2V achieve the accuracies of 90.32,62.17,65.83% and 91.59,63.27,66.63% respectively, both achieving the state-of-the-art results., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier Ltd. All rights reserved.)

For solving the facial expression recognition (FER) problem, we introduce a novel feature extractor called the coordinate-based neighborhood attention mechanism (CNAM), which uses the coordinate attention (CA) method to capture the directional relationships in separate horizontal and vertical directions, the input features from a preprocessing unit, and then passes this to two residual blocks, one consisting of the neighborhood attention (NA) mechanism, which captures the local interaction of features within the neighborhood of a feature vector, while the other one contains a channel attention implemented by a multilayer perceptron (MLP). We apply the feature extractor, the CNAM module, to four FER benchmark datasets, namely, RAF-DB, AffectNet(7cls), AffectNet(8cls), and CK+, and through qualitative and quantitative analysis techniques, we conclude that the insertion of the CNAM module could decrease the intra-cluster distances and increase the inter-cluster distances among the high-dimensional feature vectors. The CNAM compares well with other state-of-the-art (SOTA) methods, being the best-performing method for the AffectNet(7cls) and CK+ datasets, while for the RAF-DB and AffectNet(8cls) datasets, its performance is among the top-performing SOTA methods.

In studying the joint object detection and classification problem for facial expression recognition (FER) deploying the YOLOX framework, we introduce a novel feature extractor, called neighborhood coordinate attention Mamba (NCAMamba) to substitute for the original feature extractor in the Feature Pyramid Network (FPN). NCAMamba combines the background information reduction capabilities of Mamba, the local neighborhood relationship understanding of neighborhood attention, and the directional relationship understanding of coordinate attention. The resulting FER-YOLO-NCAMamba model, when applied to two unaligned FER benchmark datasets, RAF-DB and SFEW, obtains significantly improved mean average precision (mAP) scores when compared with those obtained by other state-of-the-art methods. Moreover, in ablation studies, it is found that the NCA module is relatively more important than the Visual State Space (VSS), a version of using Mamba for image processing, and in visualization studies using the grad-CAM method, it reveals that regions around the nose tip are critical to recognizing the expression; if it is too large, it may lead to erroneous prediction, while a small focused region would lead to correct recognition; this may explain why FER of unaligned faces is such a challenging problem.

In smart classroom environments, accurately recognizing students' facial expressions is crucial for teachers to efficiently assess students' learning states, timely adjust teaching strategies, and enhance teaching quality and effectiveness. In this paper, we propose a student facial expression recognition model based on multi-scale and deep fine-grained feature attention enhancement (SFER-MDFAE) to address the issues of inaccurate facial feature extraction and poor robustness of facial expression recognition in smart classroom scenarios. Firstly, we construct a novel multi-scale dual-pooling feature aggregation module to capture and fuse facial information at different scales, thereby obtaining a comprehensive representation of key facial features; secondly, we design a key region-oriented attention mechanism to focus more on the nuances of facial expressions, further enhancing the representation of multi-scale deep fine-grained feature; finally, the fusion of multi-scale and deep fine-grained attention-enhanced features is used to obtain richer and more accurate facial key information and realize accurate facial expression recognition. The experimental results demonstrate that the proposed SFER-MDFAE outperforms the existing state-of-the-art methods, achieving an accuracy of 76.18% on FER2013, 92.75% on FERPlus, 92.93% on RAF-DB, 67.86% on AffectNet, and 93.74% on the real smart classroom facial expression dataset (SCFED). These results validate the effectiveness of the proposed method.

In his 1872 monograph, Charles Darwin posited that "… the habit of expressing our feelings by certain movements, though now rendered innate, had been in some manner gradually acquired." 1 Nearly 150 years later, researchers are still teasing apart innate versus experience-dependent contributions to expression recognition. Indeed, studies have shown that face detection is surprisingly resilient to early visual deprivation, 2 , 3 , 4 , 5 pointing to plasticity that extends beyond dogmatic critical periods. 6 , 7 , 8 However, it remains unclear whether such resilience extends to downstream processing, such as the ability to recognize facial expressions. The extent to which innate versus experience-dependent mechanisms contribute to this ability has yet to be fully explored. 9 , 10 , 11 , 12 , 13 To investigate the impact of early visual experience on facial-expression recognition, we studied children with congenital cataracts who have undergone sight-correcting treatment 14 , 15 and tracked their longitudinal skill acquisition as they gain sight late in life. We introduce and explore two potential facilitators of late-life plasticity: the availability of newborn-like coarse visual acuity prior to treatment 16 and the privileged role of motion following treatment. 4 , 17 , 18 We find that early visual deprivation does not preclude partial acquisition of facial-expression recognition. While rudimentary pretreatment vision is sufficient to allow a low level of expression recognition, it does not facilitate post-treatment improvements. Additionally, only children commencing vision with high visual acuity privilege the use of dynamic cues. We conclude that skipping typical visual experience early in development and introducing high-resolution imagery late in development restricts, but does not preclude, facial-expression skill acquisition and that the representational mechanisms driving this learning differ from those that emerge during typical visual development., Competing Interests: Declaration of interests The authors declare no competing interests., (Copyright © 2024 Elsevier Inc. All rights reserved.)

In this paper, we study facial expression recognition (FER) using three modalities obtained from a light field camera: sub-aperture (SA), depth map, and all-in-focus (AiF) images. Our objective is to construct a more comprehensive and effective FER system by investigating multimodal fusion strategies. For this purpose, we employ EfficientNetV2-S, pre-trained on AffectNet, as our primary convolutional neural network. This model, combined with a BiGRU, is used to process SA images. We evaluate various fusion techniques at both decision and feature levels to assess their effectiveness in enhancing FER accuracy. Our findings show that the model using SA images surpasses state-of-the-art performance, achieving 88.13% ± 7.42% accuracy under the subject-specific evaluation protocol and 91.88% ± 3.25% under the subject-independent evaluation protocol. These results highlight our model's potential in enhancing FER accuracy and robustness, outperforming existing methods. Furthermore, our multimodal fusion approach, integrating SA, AiF, and depth images, demonstrates substantial improvements over unimodal models. The decision-level fusion strategy, particularly using average weights, proved most effective, achieving 90.13% ± 4.95% accuracy under the subject-specific evaluation protocol and 93.33% ± 4.92% under the subject-independent evaluation protocol. This approach leverages the complementary strengths of each modality, resulting in a more comprehensive and accurate FER system.

In recent years, significant progress has been made in facial expression recognition methods. However, tasks related to facial expression recognition in real environments still require further research. This paper proposes a tri-cross-attention transformer with a multi-feature fusion network (TriCAFFNet) to improve facial expression recognition performance under challenging conditions. By combining LBP (Local Binary Pattern) features, HOG (Histogram of Oriented Gradients) features, landmark features, and CNN (convolutional neural network) features from facial images, the model is provided with a rich input to improve its ability to discern subtle differences between images. Additionally, tri-cross-attention blocks are designed to facilitate information exchange between different features, enabling mutual guidance among different features to capture salient attention. Extensive experiments on several widely used datasets show that our TriCAFFNet achieves the SOTA performance on RAF-DB with 92.17%, AffectNet (7 cls) with 67.40%, and AffectNet (8 cls) with 63.49%, respectively.

Convolutional neural networks (CNNs) have made significant progress in the field of facial expression recognition (FER). However, due to challenges such as occlusion, lighting variations, and changes in head pose, facial expression recognition in real-world environments remains highly challenging. At the same time, methods solely based on CNN heavily rely on local spatial features, lack global information, and struggle to balance the relationship between computational complexity and recognition accuracy. Consequently, the CNN-based models still fall short in their ability to address FER adequately. To address these issues, we propose a lightweight facial expression recognition method based on a hybrid vision transformer. This method captures multi-scale facial features through an improved attention module, achieving richer feature integration, enhancing the network's perception of key facial expression regions, and improving feature extraction capabilities. Additionally, to further enhance the model's performance, we have designed the patch dropping (PD) module. This module aims to emulate the attention allocation mechanism of the human visual system for local features, guiding the network to focus on the most discriminative features, reducing the influence of irrelevant features, and intuitively lowering computational costs. Extensive experiments demonstrate that our approach significantly outperforms other methods, achieving an accuracy of 86.51% on RAF-DB and nearly 70% on FER2013, with a model size of only 3.64 MB. These results demonstrate that our method provides a new perspective for the field of facial expression recognition., Competing Interests: The authors declare no conflicts of interest.

Attention deficit and hyperactive disorder (ADHD) is a chronic neurodevelopmental disorder characterized by inattention, hyperactivity-impulsivity, and working memory deficits. Social dysfunction is one of the major challenges faced by children with ADHD. It has been found that children with ADHD can't perform as well as typically developing children on facial expression recognition (FER) tasks. Generally, children with ADHD have some difficulties in FER, while some studies suggest that they have no significant differences in accuracy of specific emotion recognition compared with typically developing children. The neuropsychological mechanisms underlying these difficulties are as follows. First, neuroanatomically. Compared to typically developing children, children with ADHD show smaller gray matter volume and surface area in the amygdala and medial prefrontal cortex regions, as well as reduced density and volume of axons/cells in certain frontal white matter fiber tracts. Second, neurophysiologically. Children with ADHD exhibit increased slow-wave activity in their electroencephalogram, and event-related potential studies reveal abnormalities in emotional regulation and responses to angry faces when facing facial stimuli. Third, psychologically. Psychosocial stressors may influence FER abilities in children with ADHD, and sleep deprivation in ADHD children may significantly increase their recognition threshold for negative expressions such as sadness and anger. This article reviews research progress over the past three years on FER abilities of children with ADHD, analyzing the FER deficit in children with ADHD from three dimensions: neuroanatomy, neurophysiology and psychology, aiming to provide new perspectives for further research and clinical treatment of ADHD.

Background: Emotional cognitive impairment is a core phenotype of the clinical symptoms of psychiatric disorders. The ability to measure emotional cognition is useful for assessing neurodegenerative conditions and treatment responses. However, certain factors such as culture, gender, and generation influence emotional recognition, and these differences require examination. We investigated the characteristics of healthy young Japanese adults with respect to facial expression recognition., Methods: We generated 17 models of facial expressions for each of the six basic emotions (happiness, sadness, anger, fear, disgust, and surprise) at three levels of emotional intensity using the Facial Acting Coding System (FACS). Thirty healthy Japanese young adults evaluated the type of emotion and emotional intensity the models represented to them., Results: Assessment accuracy for all emotions, except fear, exceeded 60% in approximately half of the videos. Most facial expressions of fear were rarely accurately recognized. Gender differences were observed with respect to both faces and participants, indicating that expressions on female faces were more recognizable than those on male faces, and female participants had more accurate perceptions of facial emotions than males., Conclusion: The videos used may constitute a dataset, with the possible exception of those that represent fear. The subject's ability to recognize the type and intensity of emotions was affected by the gender of the portrayed face and the evaluator's gender. These gender differences must be considered when developing a scale of facial expression recognition., (© 2023. BioMed Central Ltd., part of Springer Nature.)

In our daily lives, we routinely look at the faces of others to try to understand how they are feeling. Few studies have examined the perceptual strategies that are used to recognize facial expressions of emotion, and none have attempted to isolate visual information use with eye movements throughout development. Therefore, we recorded the eye movements of children from 5 years of age up to adulthood during recognition of the six "basic emotions" to investigate when perceptual strategies for emotion recognition become mature (i.e., most adult-like). Using iMap4, we identified the eye movement fixation patterns for recognition of the six emotions across age groups in natural viewing and gaze-contingent (i.e., expanding spotlight) conditions. While univariate analyses failed to reveal significant differences in fixation patterns, more sensitive multivariate distance analyses revealed a U-shaped developmental trajectory with the eye movement strategies of the 17- to 18-year-old group most similar to adults for all expressions. A developmental dip in strategy similarity was found for each emotional expression revealing which age group had the most distinct eye movement strategy from the adult group: the 13- to 14-year-olds for sadness recognition; the 11- to 12-year-olds for fear, anger, surprise, and disgust; and the 7- to 8-year-olds for happiness. Recognition performance for happy, angry, and sad expressions did not differ significantly across age groups, but the eye movement strategies for these expressions diverged for each group. Therefore, a unique strategy was not a prerequisite for optimal recognition performance for these expressions. Our data provide novel insights into the developmental trajectories underlying facial expression recognition, a critical ability for adaptive social relations., (Copyright © 2022 The Author(s). Published by Elsevier Inc. All rights reserved.)

This paper aims to explore the potential offered by emotion recognition systems to provide a feasible response to the growing need for audience understanding and development in the field of arts organizations. Through an empirical study, it was investigated whether the emotional valence measured on the audience through an emotion recognition system based on facial expression analysis can be used with an experience audit to: (1) support the understanding of the emotional responses of customers toward any clue that characterizes a staged performance; and (2) systematically investigate the customer's overall experience in terms of their overall satisfaction. The study was carried out in the context of opera live shows in the open-air neoclassical theater Arena Sferisterio in Macerata, during 11 opera performances. A total of 132 spectators were involved. Both the emotional valence provided by the considered emotion recognition system and the quantitative data related to customers' satisfaction, collected through a survey, were considered. Results suggest how collected data can be useful for the artistic director to estimate the audience's overall level of satisfaction and make choices about the specific characteristics of the performance, and that emotional valence measured on the audience during the show can be useful to predict overall customer satisfaction, as measured using traditional self-report methods.

Recognizing facial expressions has been a persistent goal in the scientific community. Since the rise of artificial intelligence, convolutional neural networks (CNN) have become popular to recognize facial expressions, as images can be directly used as input. Current CNN models can achieve high recognition rates, but they give no clue about their reasoning process. Explainable artificial intelligence (XAI) has been developed as a means to help to interpret the results obtained by machine learning models. When dealing with images, one of the most-used XAI techniques is LIME. LIME highlights the areas of the image that contribute to a classification. As an alternative to LIME, the CEM method appeared, providing explanations in a way that is natural for human classification: besides highlighting what is sufficient to justify a classification, it also identifies what should be absent to maintain it and to distinguish it from another classification. This study presents the results of comparing LIME and CEM applied over complex images such as facial expression images. While CEM could be used to explain the results on images described with a reduced number of features, LIME would be the method of choice when dealing with images described with a huge number of features.

With a large number of Local Binary Patterns (LBP) variants being currently used today, the significant and importance of visual descriptors in computer vision applications are prominent. This paper presents a novel visual descriptor, i.e., SIM-LBP. It employs a new matrix technique called the Symmetric Inline Matrix generator method, which acts as a new variant of LBP. The key feature that separates our variant from existing counterparts is that our variant is very efficient in extracting facial expression features like eyes, eye brows, nose and mouth in a wide range of lighting conditions. For testing our model, we applied SIM-LBP on the JAFFE dataset to convert all the images to its corresponding SIM-LBP transformed variant. These transformed images are then used to train a Convolution Neural Network (CNN) based deep learning model for facial expressions recognition (FER). Several performance evaluation metrics, i.e., recognition accuracy rate, precision, recall, and F1-score, were used to test mode efficiency in comparison with those using the traditional LBP descriptor and other LBP variants. Our model outperformed in all four matrices with the proposed SIM-LBP transformation on the input images against those of baseline methods. In comparison analysis with the other state-of-the-art methods, it shows the usefulness of the proposed SIM-LBP model. Our proposed SIM-LBP variant transformation can also be applied on facial images to identify a person's mental states and predict mood variations.

Background and Objective: In daily life, face information has the characteristics of uniqueness and universality. However, in a real-world scene, the image information of the face acquired by the acquisition device often contains noises such as blurring and sharpening. As such, super-resolution reconstruction of face features recognition based on manifold learning is proposed in this paper., Methods: We reconstruct low-resolution facial expression images, introduce a simplified residual block network and manifold learning, and propose joint supervision through a new hybrid loss function, which not only retains the color and characteristics of the image, but also retains the high-frequency information. The ResNet50 network uses the weight feature of information entropy to optimize the information of the pooling layer, and the esNet50 network uses the improved PSO algorithm to optimize the initial weight of the error back-propagation phase., Results: In the case of inputting extremely low resolution (6 × 6) facial expression images, the accuracy rate is increased by 9.091%. The accuracy of the high-resolution facial expressions after reconstruction with a size of 12×12 is 96.970%. The accuracy rate for happy expressions is 100%, the accuracy rate for anger, disgust, sadness, and surprise recognition is 97%, the accuracy rate for contempt is 94%, and the accuracy rate for fear is 88%., Conclusions: The experimental results verify the feasibility and superiority of the system, and effectively improve the accuracy of low-resolution facial expressions., Competing Interests: Declaration of Competing Interest The authors declare that there is no conflict of interests., (Copyright © 2022. Published by Elsevier B.V.)

Facial motion analysis is a research field with many practical applications, and has been strongly developed in the last years. However, most effort has been focused on the recognition of basic facial expressions of emotion and neglects the analysis of facial motions related to non-verbal communication signals. This paper focuses on the classification of facial expressions that are of the utmost importance in sign languages (Grammatical Facial Expressions) but also present in expressive spoken language. We have collected a dataset of Spanish Sign Language sentences and extracted the intervals for three types of Grammatical Facial Expressions: negation, closed queries and open queries. A study of several deep learning models using different input features on the collected dataset (LSE_GFE) and an external dataset (BUHMAP) shows that GFEs can be learned reliably with Graph Convolutional Networks simply fed with face landmarks.

Automatic facial expression recognition is essential for many potential applications. Thus, having a clear overview on existing datasets that have been investigated within the framework of face expression recognition is of paramount importance in designing and evaluating effective solutions, notably for neural networks-based training. In this survey, we provide a review of more than eighty facial expression datasets, while taking into account both macro- and micro-expressions. The proposed study is mostly focused on spontaneous and in-the-wild datasets, given the common trend in the research is that of considering contexts where expressions are shown in a spontaneous way and in a real context. We have also provided instances of potential applications of the investigated datasets, while putting into evidence their pros and cons. The proposed survey can help researchers to have a better understanding of the characteristics of the existing datasets, thus facilitating the choice of the data that best suits the particular context of their application.

Objective: People can be categorized into one of four meaning-in-life profiles: High Presence High Search (HPHS), High Presence Low Search (HPLS), Low Presence High Search (LPHS), and Low Presence Low Search (LPLS).The main goal of this study is to provide a theoretical explanation for why Chinese people with different meaning-in-life profiles have different mental health levels than Western people, based on their emotional-cognitive-processing ability., Method: We adopted eye-movement analysis and recognition-judgment experimental paradigm concerning absolute-recognition judgment and relative-recognition judgment in our study. Moreover, we applied a multifactor and multilevel mixed-experimental design. We selected 118 participants for the experiments from the 788 Chinese college students who responded., Results: Our results showed that HPHS individuals preferred positive-emotion pictures, LPLS individuals preferred negative-emotion pictures, HPLS individuals preferred positive- and neutral-emotion pictures, and LPHS individuals preferred neutral-emotion pictures. Moreover, HPHS individuals were better at accurately processing facial expression from pictures, while LPLS individuals lacked such ability. The fine-processing ability of HPLS and LPHS individuals was lower than that of HPHS yet higher than that of LPLS individuals. Moreover, the features of HPLS individuals were closer to HPHS, while those of LPHS individuals were closer to LPLS., Conclusions: Our findings support the hypothesis that meaning-in-life profiles have different immediate processing abilities and preferences regarding facial expression recognition and different emotional-cognitive-processing ability., (© 2020 Wiley Periodicals LLC.)

Purpose: Imaging, breath-holding/gating, and fixation devices have been developed to minimize setup errors so that the prescribed dose can be exactly delivered to the target volume in radiotherapy. Despite these efforts, additional patient monitoring devices have been installed in the treatment room to view patients' whole-body movement. We developed a facial expression recognition system using deep learning with a convolutional neural network (CNN) to predict patients' advanced movement, enhancing the stability of the radiation treatment by giving warning signs to radiation therapists., Materials and Methods: Convolutional neural network model and extended Cohn-Kanade datasets with 447 facial expressions of source images for training were used. Additionally, a user interface that can be used in the treatment control room was developed to monitor real-time patient's facial expression in the treatment room, and the entire system was constructed by installing a camera in the treatment room. To predict the possibility of patients' sudden movement, we categorized facial expressions into two groups: (a) uncomfortable expressions and (b) comfortable expressions. We assumed that the warning sign about the sudden movement was given when the uncomfortable expression was recognized., Results: We have constructed the facial expression monitoring system, and the training and test accuracy were 100% and 85.6%, respectively. In 10 patients, their emotions were recognized based on their comfortable and uncomfortable expressions with 100% detection rate. The detected various emotions were represented by a heatmap and motion prediction accuracy was analyzed for each patient., Conclusion: We developed a system that monitors the patient's facial expressions and predicts patient's advanced movement during the treatment. It was confirmed that our patient monitoring system can be complementarily used with the existing monitoring system. This system will help in maintaining the initial setup and improving the accuracy of radiotherapy for the patients using deep learning in radiotherapy., (© 2020 The Authors. Journal of Applied Clinical Medical Physics published by Wiley Periodicals, Inc. on behalf of American Association of Physicists in Medicine.)

Background: Neuroimaging studies suggest that facial expression recognition is processed in the bilateral posterior superior temporal sulcus (pSTS). Our recent repetitive transcranial magnetic stimulation (rTMS) study demonstrates that the bilateral pSTS is causally involved in expression recognition, although involvement of the right pSTS is greater than involvement of the left pSTS., Objective: /Hypothesis: In this study, we used a dual-site TMS to investigate whether the left pSTS is functionally connected to the right pSTS during expression recognition. We predicted that if this connection exists, simultaneous TMS disruption of the bilateral pSTS would impair expression recognition to a greater extent than unilateral stimulation of the right pSTS alone., Methods: Participants attended two TMS sessions. In Session 1, participants performed an expression recognition task while rTMS was delivered to the face-sensitive right pSTS (experimental site), object-sensitive right lateral occipital complex (control site) or no rTMS was delivered (behavioural control). In Session 2, the same experimental design was used, except that continuous theta-burst stimulation (cTBS) was delivered to the left pSTS immediately before behavioural testing commenced. Session order was counter-balanced across participants., Results: In Session 1, rTMS to the rpSTS impaired performance accuracy compared to the control conditions. Crucially in Session 2, the size of this impairment effect doubled after cTBS was delivered to the left pSTS., Conclusions: Our results provide evidence for a causal functional connection between the left and right pSTS during expression recognition. In addition, this study further demonstrates the utility of the dual-site TMS for investigating causal functional links between brain regions., Competing Interests: Declaration of competing interest None., (Copyright © 2020 The Authors. Published by Elsevier Inc. All rights reserved.)

Facial expressions are one of the important non-verbal ways used to understand human emotions during communication. Thus, acquiring and reproducing facial expressions is helpful in analyzing human emotional states. However, owing to complex and subtle facial muscle movements, facial expression modeling from images with face poses is difficult to achieve. To handle this issue, we present a method for acquiring facial expressions from a non-frontal single photograph using a 3D-aided approach. In addition, we propose a contour-fitting method that improves the modeling accuracy by automatically rearranging 3D contour landmarks corresponding to fixed 2D image landmarks. The acquired facial expression input can be parametrically manipulated to create various facial expressions through a blendshape or expression transfer based on the FACS (Facial Action Coding System). To achieve a realistic facial expression synthesis, we propose an exemplar-texture wrinkle synthesis method that extracts and synthesizes appropriate expression wrinkles according to the target expression. To do so, we constructed a wrinkle table of various facial expressions from 400 people. As one of the applications, we proved that the expression-pose synthesis method is suitable for expression-invariant face recognition through a quantitative evaluation, and showed the effectiveness based on a qualitative evaluation. We expect our system to be a benefit to various fields such as face recognition, HCI, and data augmentation for deep learning.

Sound perception studies mostly depend on questionnaires with fixed indicators. Therefore, it is desirable to explore methods with dynamic outputs. The present study aims to explore the effects of sound perception in the urban environment on facial expressions using a software named FaceReader based on facial expression recognition (FER). The experiment involved three typical urban sound recordings, namely, traffic noise, natural sound, and community sound. A questionnaire on the evaluation of sound perception was also used, for comparison. The results show that, first, FER is an effective tool for sound perception research, since it is capable of detecting differences in participants' reactions to different sounds and how their facial expressions change over time in response to those sounds, with mean difference of valence between recordings from 0.019 to 0.059 (p < 0.05or p < 0.01). In a natural sound environment, for example, facial expression increased by 0.04 in the first 15 s and then went down steadily at 0.004 every 20 s. Second, the expression indices, namely, happy, sad, and surprised, change significantly under the effect of sound perception. In the traffic sound environment, for example, happy decreased by 0.012, sad increased by 0.032, and surprised decreased by 0.018. Furthermore, social characteristics such as distance from living place to natural environment (r = 0.313), inclination to communicate (r = 0.253), and preference for crowd (r = 0.296) have effects on facial expression. Finally, the comparison of FER and questionnaire survey results showed that in the traffic noise recording, valence in the first 20 s best represents acoustic comfort and eventfulness; for natural sound, valence in the first 40 s best represents pleasantness; and for community sound, valence in the first 20 s of the recording best represents acoustic comfort, subjective loudness, and calmness., Competing Interests: Declaration of competing interest We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled, “On the Effectiveness of Facial Expression Recognition for Evaluation of Urban Sound Perception”, Authors: Qi Meng, Xuejun Hu, Jian Kang, and Yue Wu., (Copyright © 2018 Elsevier B.V. All rights reserved.)

Background: Expression in H-sapiens plays a remarkable role when it comes to social communication. The identification of this expression by human beings is relatively easy and accurate. However, achieving the same result in 3D by machine remains a challenge in computer vision. This is due to the current challenges facing facial data acquisition in 3D; such as lack of homology and complex mathematical analysis for facial point digitization. This study proposes facial expression recognition in human with the application of Multi-points Warping for 3D facial landmark by building a template mesh as a reference object. This template mesh is thereby applied to each of the target mesh on Stirling/ESRC and Bosphorus datasets. The semi-landmarks are allowed to slide along tangents to the curves and surfaces until the bending energy between a template and a target form is minimal and localization error is assessed using Procrustes ANOVA. By using Principal Component Analysis (PCA) for feature selection, classification is done using Linear Discriminant Analysis (LDA)., Result: The localization error is validated on the two datasets with superior performance over the state-of-the-art methods and variation in the expression is visualized using Principal Components (PCs). The deformations show various expression regions in the faces. The results indicate that Sad expression has the lowest recognition accuracy on both datasets. The classifier achieved a recognition accuracy of 99.58 and 99.32% on Stirling/ESRC and Bosphorus, respectively., Conclusion: The results demonstrate that the method is robust and in agreement with the state-of-the-art results.

To improve the robustness and to reduce the energy consumption of facial expression recognition, this study proposed a facial expression recognition method based on improved deep residual networks (ResNets). Residual learning has solved the degradation problem of deep Convolutional Neural Networks (CNNs); therefore, in theory, a ResNet can consist of infinite number of neural layers. On the one hand, ResNets benefit from better performance on artificial intelligence (AI) tasks, thanks to its deeper network structure; meanwhile, on the other hand, it faces a severe problem of energy consumption, especially on mobile devices. Hence, this study employs a novel activation function, the Noisy Softplus (NSP), to replace rectified linear units (ReLU) to get improved ResNets. NSP is a biologically plausible activation function, which was first proposed in training Spiking Neural Networks (SNNs); thus, NSP-trained models can be directly implemented on ultra-low-power neuromorphic hardware. We built an 18-layered ResNet using NSP to perform facial expression recognition across datasets Cohn-Kanade (CK+), Karolinska Directed Emotional Faces (KDEF) and GENKI-4K. The results achieved better anti-noise ability than ResNet using the activation function ReLU and showed low energy consumption running on neuromorphic hardware. This study not only contributes a solution for robust facial expression recognition, but also consolidates the low energy cost of their implementation on neuromorphic devices, which could pave the way for high-performance, noise-robust and energy-efficient vision applications on mobile hardware.

Child-Robot Interaction (CRI) has become increasingly addressed in research and applications. This work proposes a system for emotion recognition in children, recording facial images by both visual (RGB-red, green and blue) and Infrared Thermal Imaging (IRTI) cameras. For this purpose, the Viola-Jones algorithm is used on color images to detect facial regions of interest (ROIs), which are transferred to the thermal camera plane by multiplying a homography matrix obtained through the calibration process of the camera system. As a novelty, we propose to compute the error probability for each ROI located over thermal images, using a reference frame manually marked by a trained expert, in order to choose that ROI better placed according to the expert criteria. Then, this selected ROI is used to relocate the other ROIs, increasing the concordance with respect to the reference manual annotations. Afterwards, other methods for feature extraction, dimensionality reduction through Principal Component Analysis (PCA) and pattern classification by Linear Discriminant Analysis (LDA) are applied to infer emotions. The results show that our approach for ROI locations may track facial landmarks with significant low errors with respect to the traditional Viola-Jones algorithm. These ROIs have shown to be relevant for recognition of five emotions, specifically disgust, fear, happiness, sadness, and surprise, with our recognition system based on PCA and LDA achieving mean accuracy (ACC) and Kappa values of 85.75% and 81.84%, respectively. As a second stage, the proposed recognition system was trained with a dataset of thermal images, collected on 28 typically developing children, in order to infer one of five basic emotions (disgust, fear, happiness, sadness, and surprise) during a child-robot interaction. The results show that our system can be integrated to a social robot to infer child emotions during a child-robot interaction.

In recent years, researchers of deep neural networks (DNNs)-based facial expression recognition (FER) have reported results showing that these approaches overcome the limitations of conventional machine learning-based FER approaches. However, as DNN-based FER approaches require an excessive amount of memory and incur high processing costs, their application in various fields is very limited and depends on the hardware specifications. In this paper, we propose a fast FER algorithm for monitoring a driver's emotions that is capable of operating in low specification devices installed in vehicles. For this purpose, a hierarchical weighted random forest (WRF) classifier that is trained based on the similarity of sample data, in order to improve its accuracy, is employed. In the first step, facial landmarks are detected from input images and geometric features are extracted, considering the spatial position between landmarks. These feature vectors are then implemented in the proposed hierarchical WRF classifier to classify facial expressions. Our method was evaluated experimentally using three databases, extended Cohn-Kanade database (CK+), MMI and the Keimyung University Facial Expression of Drivers (KMU-FED) database, and its performance was compared with that of state-of-the-art methods. The results show that our proposed method yields a performance similar to that of deep learning FER methods as 92.6% for CK+ and 76.7% for MMI, with a significantly reduced processing cost approximately 3731 times less than that of the DNN method. These results confirm that the proposed method is optimized for real-time embedded applications having limited computing resources., Competing Interests: The authors declare no conflict of interest.

Previous research showed that we use different regions of a face to categorize different facial expressions, e.g. mouth region for identifying happy faces; eyebrows, eyes and upper part of nose for identifying angry faces. These findings imply that the spatial information along or close to the horizontal orientation might be more useful than others for facial expression recognition. In this study, we examined how the performance for recognizing facial expression depends on the spatial information along different orientations, and whether the pixel-level differences in the face images could account for subjects' performance. Four facial expressions-angry, fearful, happy and sad-were tested. An orientation filter (bandwidth = 23°) was applied to restrict information within the face images, with the center of the filter ranged from 0° (horizontal) to 150° in steps of 30°. Accuracy for recognizing facial expression was measured for an unfiltered and the six filtered conditions. For all four facial expressions, recognition performance (normalized d') was virtually identical for filter orientations of -30°, horizontal and 30°, and declined systematically as the filter orientation approached vertical. The information contained in mouth and eye regions is a significant predictor for subject's response (based on the confusion patterns). We conclude that young adults with normal vision categorizes facial expression most effectively based on the spatial information around the horizontal orientation which captures primary changes of facial features across expressions. Across all spatial orientations, the information contained in mouth and eye regions contributes significantly to facial expression categorization., (Copyright © 2018 Elsevier Ltd. All rights reserved.)

Background: The facial expression of medical staff has been known to greatly affect the psychological state of patients, making them feel uneasy or conversely, cheering them up. By clarifying the characteristics of facial expression recognition ability in patients with Lewy body disease, the aim of this study is to examine points to facilitate smooth communication between caregivers and patients with the disease whose cognitive function has deteriorated., Methods: During the period from March 2016 to July 2017, we examined the characteristics of recognition of the six facial expressions of "happiness," "sadness," "fear," "anger," "surprise," and "disgust" for 107 people aged 60 years or more, both outpatient and inpatient, who hospital specialists had diagnosed with Lewy body diseases of Parkinson's disease, Parkinson's disease with dementia, and dementia with Lewy bodies. Based on facial expression recognition test results, we classified them by cluster analysis and clarified features of each type., Results: In patients with Lewy body disease, happiness was kept unaffected by aging, age of onset, duration of the disease, cognitive function, and apathy; however, recognizing the facial expression of fear was difficult. In addition, due to aging, cognitive decline, and apathy, the facial expression recognition ability for sadness and anger decreased. In particular, cognitive decline reduced recognition of all of the facial expressions except for happiness. The test accuracy rates were classified into three types using the cluster analysis: "stable type," "mixed type," and "reduced type". In the "reduced type", the overall facial recognition ability declined except happiness, and in the mixed type, recognition ability of anger particularly declined., Conclusion: There were several facial expressions that the Lewy body disease patients were unable to accurately identify. Caregivers are recommended to make an effort to compensate for such situations with language or body contact, etc., as a way to convey correct feeling to the patients of each type.

Introduction: The aim of this study was to investigate whether social anxiety disorder (SAD) patients have low emotion recognition accuracy, take longer for emotion recognition and tended to interpret a stimulus more negatively than controls., Methods: Fifty-six SAD patients and 56 healthy controls were participated in this study. We evaluated facial emotion recognition using a computer program in which participants chose one of seven emotions as displayed in each of 55 photographs of faces. We compared the accuracy and reaction times of the patients and controls. We further analysed which emotions were selected in the incorrect answers., Results: SAD patients showed delayed reaction times for all emotions except anger and lower accuracy for fear, surprise, neutrality and happiness compared to controls. After applying Bonferroni correction, only delayed reaction time for surprise and happiness were remain significant. Neutrality was not mistaken for a negative emotion at a higher rate by SAD patients than by controls., Conclusion: This result suggests that the alterations of reaction time and accuracy of emotion recognition of SAD patients, especially in emotions with positive valence, play a more important role than negative bias in the cognitive aspects of SAD.

As a simulation of human expression recognition, the studies on automatic expression recognition expect to draw useful enlightenment through close, accurate observation on human expression processing via advanced devices. Eye-trackers are mostly used devices that are technically designed to obtain eye-movement data. However, due to the discrepancy between target faces, across-target analysis is limited in these studies, and this much reduces the chance of finding the latent eye-behavior patterns. Through the utilization of correspondences between targets, this study achieves an across-target analysis to explore the attention pattern in expression recognition. The fixations from different targets are mapped onto a synthetic face to generate an across-target fixation map, and then tokenized with area of interests (AOI), measured in receiver operating characteristic (ROC) space, modeled by linear regression and compared through Pearson's correlation. The resulted averaged correlation values vary in the range (0.60, 0.86), and illustrate that there is significant similarity between subjects when recognizing the same expression classes.

In this study, it is aimed to determine the degree of the development in emotional expression of full face transplant patients from photographs. Hence, a rehabilitation process can be planned according to the determination of degrees as a later work. As envisaged, in full face transplant cases, the determination of expressions can be confused or cannot be achieved as the healthy control group. In order to perform image-based analysis, a control group consist of 9 healthy males and 2 full-face transplant patients participated in the study. Appearance-based Gabor Wavelet Transform (GWT) and Local Binary Pattern (LBP) methods are adopted for recognizing neutral and 6 emotional expressions which consist of angry, scared, happy, hate, confused and sad. Feature extraction was carried out by using both methods and combination of these methods serially. In the performed expressions, the extracted features of the most distinct zones in the facial area where the eye and mouth region, have been used to classify the emotions. Also, the combination of these region features has been used to improve classifier performance. Control subjects and transplant patients' ability to perform emotional expressions have been determined with K-nearest neighbor (KNN) classifier with region-specific and method-specific decision stages. The results have been compared with healthy group. It has been observed that transplant patients don't reflect some emotional expressions. Also, there were confusions among expressions.

Facial emotion perception is a fundamental social competency relying on a specialised, yet distributed, neural network. This review aimed to determine whether patients with epilepsy have facial emotion perception accuracy impairments overall, or for a subset of emotions (anger, disgust, happiness, sadness, fear, and surprise), and the relationship to epilepsy type, demographic/treatment variables, and brain organisation. Database searches used PRISMA guidelines with strict inclusion/exclusion criteria. Thirty included studies assessed patients with temporal lobe (TLE; n=709), frontocentral (FCE; n=22), and genetic generalised (GGE; n=48) epilepsy. Large deficits emerged in patients with epilepsy compared to controls (n=746; Hedges' g=0.908-1.076). Patients with TLE were significantly impaired on all emotions except surprise; patients with GGE were significantly impaired in anger, disgust, and fear perception. Meta-regression of patients with TLE revealed younger age at testing was associated with lower accuracy. This review provides evidence for marked global deficits of emotion perception in epilepsy, with differential emotion-specific impairment patterns in patients with TLE and GGE., (Crown Copyright © 2017. Published by Elsevier Ltd. All rights reserved.)

Studies have examined social maladjustment among individuals with Internet addiction, but little is known about their deficits in specific social skills and the underlying psychological mechanisms. The present study filled these gaps by (a) establishing a relationship between deficits in facial expression recognition and Internet addiction, and (b) examining the mediating role of perceived stress that explains this hypothesized relationship. Ninety-seven participants completed validated questionnaires that assessed their levels of Internet addiction and perceived stress, and performed a computer-based task that measured their facial expression recognition. The results revealed a positive relationship between deficits in recognizing disgust facial expression and Internet addiction, and this relationship was mediated by perceived stress. However, the same findings did not apply to other facial expressions. Ad hoc analyses showed that recognizing disgust was more difficult than recognizing other facial expressions, reflecting that the former task assesses a social skill that requires cognitive astuteness. The present findings contribute to the literature by identifying a specific social skill deficit related to Internet addiction and by unveiling a psychological mechanism that explains this relationship, thus providing more concrete guidelines for practitioners to strengthen specific social skills that mitigate both perceived stress and Internet addiction., (Copyright © 2017 Elsevier Ireland Ltd. All rights reserved.)

In the pattern recognition domain, deep architectures are currently widely used and they have achieved fine results. However, these deep architectures make particular demands, especially in terms of their requirement for big datasets and GPU. Aiming to gain better results without deep networks, we propose a simplified algorithm framework using fusion features extracted from the salient areas of faces. Furthermore, the proposed algorithm has achieved a better result than some deep architectures. For extracting more effective features, this paper firstly defines the salient areas on the faces. This paper normalizes the salient areas of the same location in the faces to the same size; therefore, it can extracts more similar features from different subjects. LBP and HOG features are extracted from the salient areas, fusion features' dimensions are reduced by Principal Component Analysis (PCA) and we apply several classifiers to classify the six basic expressions at once. This paper proposes a salient areas definitude method which uses peak expressions frames compared with neutral faces. This paper also proposes and applies the idea of normalizing the salient areas to align the specific areas which express the different expressions. As a result, the salient areas found from different subjects are the same size. In addition, the gamma correction method is firstly applied on LBP features in our algorithm framework which improves our recognition rates significantly. By applying this algorithm framework, our research has gained state-of-the-art performances on CK+ database and JAFFE database.

The recognition of facial expressions of emotion is adaptive for human social interaction, but the ability to do this and the manner in which it is achieved differs among individuals. Previous functional neuroimaging studies have demonstrated that some brain regions, such as the inferior frontal gyrus (IFG), are active during the response to emotional facial expressions in healthy participants, and lesion studies have demonstrated that damage to these structures impairs the recognition of facial expressions. However, it remains to be established whether individual differences in the structure of these regions could be associated with differences in the ability to recognize facial expressions. We investigated this issue using acquired structural magnetic resonance imaging, and assessed the performance of healthy adults with respect to recognition of the facial expressions of six basic emotions. The gray matter volume of the right IFG positively correlated with the total accuracy of facial expression recognition. This suggests that individual differences in the ability to recognize facial expressions are associated with differences in the structure of the right IFG. Furthermore, the mirror neuron activity of the IFG may be important for establishing efficient facial mimicry to facilitate emotion recognition., (© The Author (2016). Published by Oxford University Press. For Permissions, please email: journals.permissions@oup.com.)

Face perception is essential for daily and social activities. Neuroimaging studies have revealed a distributed face network (FN) consisting of multiple regions that exhibit preferential responses to invariant or changeable facial information. However, our understanding about how these regions work collaboratively to facilitate facial information processing is limited. Here, we focused on changeable facial information processing, and investigated how the functional integration of the FN is related to the performance of facial expression recognition. To do so, we first defined the FN as voxels that responded more strongly to faces than objects, and then used a voxel-based global brain connectivity method based on resting-state fMRI to characterize the within-network connectivity (WNC) of each voxel in the FN. By relating the WNC and performance in the "Reading the Mind in the Eyes" Test across participants, we found that individuals with stronger WNC in the right posterior superior temporal sulcus (rpSTS) were better at recognizing facial expressions. Further, the resting-state functional connectivity (FC) between the rpSTS and right occipital face area (rOFA), early visual cortex (EVC), and bilateral STS were positively correlated with the ability of facial expression recognition, and the FCs of EVC-pSTS and OFA-pSTS contributed independently to facial expression recognition. In short, our study highlights the behavioral significance of intrinsic functional integration of the FN in facial expression processing, and provides evidence for the hub-like role of the rpSTS for facial expression recognition. Hum Brain Mapp 37:1930-1940, 2016. © 2016 Wiley Periodicals, Inc., (© 2016 Wiley Periodicals, Inc.)

The orbitofrontal cortex (OFC) has been implicated in the capacity to accurately recognise facial expressions. The aim of the current study was to determine if anodal transcranial direct current stimulation (tDCS) targeting the right OFC in healthy adults would enhance facial expression recognition, compared with a sham condition. Across two counterbalanced sessions of tDCS (i.e. anodal and sham), 20 undergraduate participants (18 female) completed a facial expression labelling task comprising angry, disgusted, fearful, happy, sad and neutral expressions, and a control (social judgement) task comprising the same expressions. Responses on the labelling task were scored for accuracy, median reaction time and overall efficiency (i.e. combined accuracy and reaction time). Anodal tDCS targeting the right OFC enhanced facial expression recognition, reflected in greater efficiency and speed of recognition across emotions, relative to the sham condition. In contrast, there was no effect of tDCS to responses on the control task. This is the first study to demonstrate that anodal tDCS targeting the right OFC boosts facial expression recognition. This finding provides a solid foundation for future research to examine the efficacy of this technique as a means to treat facial expression recognition deficits, particularly in individuals with OFC damage or dysfunction., (© The Author (2015). Published by Oxford University Press. For Permissions, please email: journals.permissions@oup.com.)

Unlabelled: We studied the ability of individuals with mild cognitive impairment (MCI) to process emotional facial expressions (EFEs). To date, no systematic study has addressed how variation in intensity affects recognition of the different type of EFEs in such subjects., Design: Two groups of 50 elderly subjects, 50 healthy individuals and 50 with MCI, completed a task that involved identifying 180 EFEs prepared using virtual models. Two features of the EFEs were contemplated, their valence (operationalized in six basic emotions) and five levels of intensity., Results: At all levels of intensity, elderly individuals with MCI were significantly worse at identifying each EFE than healthy subjects. Some emotions were easier to identify than others, with happiness proving to be the easiest to identify and disgust the hardest, and intensity influenced the identification of the EFEs (the stronger the intensity, the greater the number of correct identifications). Overall, elderly individuals with MCI had a poorer capacity to process EFEs, suggesting that cognitive ability modulates the processing of emotions, where features of such stimuli also seem to play a prominent role (e.g., valence and intensity). Thus, the neurological substrates involved in emotional processing appear to be affected by MCI., (Copyright © 2015 Elsevier B.V. All rights reserved.)

Previous studies of facial emotion processing in bipolar disorder (BD) have reported conflicting findings. In independently conducted studies, we investigate facial emotion labeling in euthymic and depressed BD patients using tasks with static and dynamically morphed images of different emotions displayed at different intensities. Study 1 included 38 euthymic BD patients and 28 controls. Participants completed two tasks: labeling of static images of basic facial emotions (anger, disgust, fear, happy, sad) shown at different expression intensities; the Eyes Test (Baron-Cohen, Wheelwright, Hill, Raste, & Plumb, 2001), which involves recognition of complex emotions using only the eye region of the face. Study 2 included 53 depressed BD patients and 47 controls. Participants completed two tasks: labeling of "dynamic" facial expressions of the same five basic emotions; the Emotional Hexagon test (Young, Perret, Calder, Sprengelmeyer, & Ekman, 2002). There were no significant group differences on any measures of emotion perception/labeling, compared to controls. A significant group by intensity interaction was observed in both emotion labeling tasks (euthymia and depression), although this effect did not survive the addition of measures of executive function/psychomotor speed as covariates. Only 2.6-15.8% of euthymic patients and 7.8-13.7% of depressed patients scored below the 10th percentile of the controls for total emotion recognition accuracy. There was no evidence of specific deficits in facial emotion labeling in euthymic or depressed BD patients. Methodological variations-including mood state, sample size, and the cognitive demands of the tasks-may contribute significantly to the variability in findings between studies.

Our objective was to compare the ability to discriminate and categorize emotional facial expressions (EFEs) and facial identity characteristics (age and/or gender) in a group of 53 individuals with Parkinson's disease (PD) and another group of 53 healthy subjects. On the one hand, by means of discrimination and identification tasks, we compared two stages in the visual recognition process that could be selectively affected in individuals with PD. On the other hand, facial expression versus gender and age comparison permits us to contrast whether the emotional or non-emotional content influences the configural perception of faces. In Experiment I, we did not find differences between groups, either with facial expression or age, in discrimination tasks. Conversely, in Experiment II, we found differences between the groups, but only in the EFE identification task. Taken together, our results indicate that configural perception of faces does not seem to be globally impaired in PD. However, this ability is selectively altered when the categorization of emotional faces is required. A deeper assessment of the PD group indicated that decline in facial expression categorization is more evident in a subgroup of patients with higher global impairment (motor and cognitive). Taken together, these results suggest that the problems found in facial expression recognition may be associated with the progressive neuronal loss in frontostriatal and mesolimbic circuits, which characterizes PD., (© 2013 The British Psychological Society.)

We cross-sectionally examined the development of the ability to recognize facial expressions of emotions in preschool-aged children and the relationship between this ability and other neurocognitive capacities, that is, attention/executive functions, language, memory/learning, sensorimotor functions, theory of mind, and visuospatial processing. Children aged 3 to 6 years with no significant developmental deficits (N = 370) were assessed with a nonverbal matching task of emotion recognition ability: The Affect Recognition subtest from the NEPSY-II. The relationship between emotion recognition ability and other neurocognitive capacities was analyzed using correlation, regression, and commonality analyses. The results showed that (a) emotion recognition ability improved with age-this development decelerating mildly between ages 5 and 6-(b) emotion recognition ability correlated with all other neurocognitive capacities, and (c) language, attention/executive functions, and theory of mind were significant predictors of emotion recognition ability in the regression analysis. As revealed by the commonality analysis, and in contrast to most previous studies, language was the most important predictor of nonverbal emotion recognition ability. These results suggest that nonverbal emotion matching is an early maturing skill that develops in relation to other neurocognitive capacities, especially linguistic ability.

Facial expression recognition is an interesting and challenging subject. Considering the nonlinear manifold structure of facial images, a new kernel-based manifold learning method, called kernel discriminant isometric mapping (KDIsomap), is proposed. KDIsomap aims to nonlinearly extract the discriminant information by maximizing the interclass scatter while minimizing the intraclass scatter in a reproducing kernel Hilbert space. KDIsomap is used to perform nonlinear dimensionality reduction on the extracted local binary patterns (LBP) facial features, and produce low-dimensional discrimimant embedded data representations with striking performance improvement on facial expression recognition tasks. The nearest neighbor classifier with the Euclidean metric is used for facial expression classification. Facial expression recognition experiments are performed on two popular facial expression databases, i.e., the JAFFE database and the Cohn-Kanade database. Experimental results indicate that KDIsomap obtains the best accuracy of 81.59% on the JAFFE database, and 94.88% on the Cohn-Kanade database. KDIsomap outperforms the other used methods such as principal component analysis (PCA), linear discriminant analysis (LDA), kernel principal component analysis (KPCA), kernel linear discriminant analysis (KLDA) as well as kernel isometric mapping (KIsomap).

Face expression recognition in the wild faces challenges such as small data size, low quality images, and noisy labels. In order to solve these problems, this paper proposes a novel structural self-contrast learning method (SSCL) based on adaptively weighted negative samples. Firstly, two augmented variants are generated for each input sample as positive samples, which should be as similar in structure as possible. Secondly, EMD (Earth's Mover Distance) is used to define the structural similarity between the two augmented samples and extract the features in a self-supervised way. Subsequently, the concept of negative samples is further introduced into SSCL, and the augmented variants of other categories are generated as negative samples, which makes the structural similarity between positive samples increase and between negative samples simultaneously decrease. Finally, an adaptive weighting method for the negative is proposed, which aims to solve the imbalance between positive samples and negative samples, mine difficult negative samples, and reduce their similarity with the input image in an adaptive way. The extensive experiments on benchmark datasets demonstrate that SSCL outperforms the state-of-the-art methods. The cross-dataset evaluation shows the superior generalization capability of the proposed method. The code will be available at https://github.com/zjhzhy/SSCL. [ABSTRACT FROM AUTHOR]

In the field of facial expression recognition (FER), two main trends point to the data-driven FER and feature-driven FER exist. The former focused on the data problems (e.g., sample imbalance and multimodal fusion), while the latter explored the facial expression features. As the feature-driven FER is more important than the data-driven FER, for deeper mining of facial features, we propose an expression recognition model based on Local–Global information Reasoning and Landmark Spatial Distributions. Particularly to reason local–global information, multiple attention mechanisms with the modified residual module are designed for the Res18-LG module. In addition, taking the spatial topology of facial landmarks into account, a topological relationship graph of landmarks and a two-layer graph neural network are introduced to extract spatial distribution features. Finally, the experiment results on FERPlus and RAF-DB datasets demonstrate that our model outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

This work introduces the representation ensemble learning algorithm, a novel approach for generating diverse unsupervised representations rooted in the principles of self-taught learning. The ensemble comprises convolutional autoencoders (CAEs) learned in an unsupervised manner, fostering diversity via a loss function designed to penalize similar CAEs' latent representations. We employ support vector machines, bagging, and random forest as primary classification methods for the final classification step. Additionally, we incorporate KnoraU, a well-established technique used to dynamically select competent classifiers based on a test sample. We evaluate various fusion strategies, including sum, product, and stacking, to comprehensively assess the ensemble's performance. A robust experimental protocol considering the facial expression recognition problem shows that the proposed approach based on self-taught learning surpasses the accuracy of fine-tuned convolutional neural network (CNN) models. In terms of accuracy, the proposed method is up to 9.9 and 6.3 percentage points better than the CNN-based models fine-tuned for JAFFE and CK+ datasets, respectively. [ABSTRACT FROM AUTHOR]

The cognitive emotions of individuals during tasks largely determine the success or failure of tasks in various fields such as the military, medical, industrial fields, etc. Facial video data can carry more emotional information than static images because emotional expression is a temporal process. Video-based Facial Expression Recognition (FER) has received increasing attention from the relevant scholars in recent years. However, due to the high cost of marking and training video samples, feature extraction is inefficient and ineffective, which leads to a low accuracy and poor real-time performance. In this paper, a cognitive emotion recognition method based on video data is proposed, in which 49 emotion description points were initially defined, and the spatial–temporal features of cognitive emotions were extracted from the video data through a feature extraction method that combines geodesic distances and sample entropy. Then, an active learning algorithm based on complexity and uncertainty was proposed to automatically select the most valuable samples, thereby reducing the cost of sample labeling and model training. Finally, the effectiveness, superiority, and real-time performance of the proposed method were verified utilizing the MMI Facial Expression Database and some real-time-collected data. Through comparisons and testing, the proposed method showed satisfactory real-time performance and a higher accuracy, which can effectively support the development of a real-time monitoring system for cognitive emotions. [ABSTRACT FROM AUTHOR]

Facial expression recognition (FER) is significant for daily mental health monitoring because facial expressions reflect an individual's mental condition. However, it is still hard to achieve accurate and convenient FER using wearable devices. Here, a high‐accuracy, self‐powered, and intelligent FER system is reported consisting of a triboelectric hydrogel sensor network to collect facial expression signals and a deep learning model to process and recognize the signals. The triboelectric hydrogel sensors are demonstrated to show excellent properties, such as 50% stretchability, 90% transparency, and a response time of 48 ms. With a 1D convolutional neural network, six basic expressions can be recognized with an average recognition accuracy of 99.44%. Finally, a 3D virtual character model is built on a computer to display real emotions synchronously. Compared with previous similar reports, this system can recognize more types of expressions with significantly improved accuracy. Therefore, this work can potentially not only help doctors to determine a patient's mental health condition through virtual communication while protecting the patient's privacy but also provide a highly promising approach for virtual telemedicine in the future. [ABSTRACT FROM AUTHOR]

People routinely use facial expressions to communicate successfully and to regulate other’s behaviour, yet modelling the form and meaning of these facial behaviours has proven surprisingly complex. One reason for this difficulty may lie in an over-reliance on the assumptions inherent in existing theories of facial expression – specifically that (1) there is a putative set of facial expressions that signal an internal state of emotion, (2) patterns of facial movement have been empirically linked to the prototypical emotions in this set, and (3) static, non-social, posed images from convenience samples are adequate to validate the first two assumptions. These assumptions have guided the creation of datasets, which are then used to train unrepresentative computational models of facial expression. In this article, we discuss existing theories of facial expression and review how they have shaped current facial expression recognition tools. We then discuss the resources that are available to help researchers build a more ecologically valid model of facial expressions. [ABSTRACT FROM AUTHOR]