Facial Emotion Recognition Accuracy Research Articles

Simple Histogram Equalization Technique Improves Performance of VGG Models on Facial Emotion Recognition Datasets

Facial emotion recognition (FER) is crucial across psychology, neuroscience, computer vision, and machine learning due to the diversified and subjective nature of emotions, varying considerably across individuals, cultures, and contexts. This study explored FER through convolutional neural networks (CNNs) and Histogram Equalization techniques. It investigated the impact of histogram equalization, data augmentation, and various model optimization strategies on FER accuracy across different datasets like KDEF, CK+, and FER2013. Using pre-trained VGG architectures, such as VGG19 and VGG16, this study also examined the effectiveness of fine-tuning hyperparameters and implementing different learning rate schedulers. The evaluation encompassed diverse metrics including accuracy, Area Under the Receiver Operating Characteristic Curve (AUC-ROC), Area Under the Precision–Recall Curve (AUC-PRC), and Weighted F1 score. Notably, the fine-tuned VGG architecture demonstrated a state-of-the-art performance compared to conventional transfer learning models and achieved 100%, 95.92%, and 69.65% on the CK+, KDEF, and FER2013 datasets, respectively.

Disruptive mood dysregulation disorder: the relationship between cardiac vagal tone, emotion recognition, and expressive prosody

Children with disruptive mood dysregulation disorder have difficulties with emotion regulation and recognition which impacts their social functioning. Due to its recent definition, there is limited information on the neural mechanisms of this disorder. This exploratory study examined the role of autonomic heart regulation in children with disruptive mood dysregulation disorder, motivated by suggestions of reduced autonomic regulation (measured by heart rate variability) associated with psychiatric disorders in adults and children. The cross-sectional design sampled two groups of children, one with disruptive mood dysregulation disorder ( n = 15; 14 male) and the other with typical development ( n = 15; 14 male). Heart rate variability, facial emotion recognition accuracy and speed, and prosody were measured. There were no significant differences in heart rate variability between the two groups before, or during, the tasks. However, children with disruptive mood dysregulation disorder experienced significantly greater difficulty recognising fear in faces compared to controls and confused other negatively valenced emotions. There was also a significant, negative correlation between heart rate variability and prosody modulation (verbal expression of emotion) in the control group, a relationship that was absent in the disruptive mood dysregulation disorder group. These results suggest that atypical autonomic regulation during emotionally evocative situations and difficulty with facial emotion recognition may contribute to the challenges experienced by children with disruptive mood dysregulation disorder. This is valuable knowledge for psychologists, which can contribute to individualised, effective, and comprehensive care that addresses the child’s emotional, social, and relational well-being.

Facial emotion recognition with a reduced feature set for video game and metaverse avatars

AbstractThis paper presents a novel real‐time facial feature extraction algorithm, producing a small feature set, suitable for implementing emotion recognition with online game and metaverse avatars. The algorithm aims to reduce data transmission and storage requirements, hurdles in the adoption of emotion recognition in these mediums. The early results presented show a facial emotion recognition accuracy of up to 92% on one benchmark dataset, with an overall accuracy of 77.2% across a wide range of datasets, demonstrating the early promise of the research.

Facial emotion recognition and social-emotional problems in middle childhood: Assessment of directional effects

Research has consistently shown that difficulties in facial emotion recognition (FER) are associated with peer problems and internalizing symptoms during middle childhood. However, no longitudinal research has investigated the direction of effects, that is, how these constructs influence each other across time. In this preregistered three-wave panel study, we tested the directional effects between FER, peer problems, and internalizing symptoms among Finnish school-aged children ( n = 3,607; Mage = 8.20, SDage = 0.86; 51% female). The results of random-intercept cross-lagged panel models showed that a low FER accuracy and high biases toward happiness and sadness correlated with higher levels of peer problems and internalizing symptoms at the between-person level. However, we found no evidence of directional effects at the within-person level. Overall, our findings suggest that these constructs might be associated because of shared underlying causes, rather than mutually influencing one another in middle childhood.

Voluntary imitation of dynamic facial expressions in attention deficit hyperactivity disorder: a facial-behavior analysis

ABSTRACT Objective The difficulties involved in social interaction among children with attention deficit hyperactivity disorder (ADHD) have been shown in many studies. Based on the knowledge that the imitation of facial expressions is a key factor in social interaction and functionality, the focus of prior studies has been on the evaluation of facial expressions in individuals with ADHD. However, little is known about voluntary facial mimicry in individuals with ADHD. In this context, we aimed to evaluate the voluntary-facial-imitation intensity of dynamic facial expressions in children with ADHD. Method Forty-one children with ADHD and 53 typically developing children were included in the study. Participants were presented with a video of six basic emotions and neutral facial expressions selected from the EU-Emotion Stimulus Set via a screen. After each emotion, the instruction “now imitate it” was given. While the children watched the video, their faces were recorded with a webcam. The intensity of the children’s voluntary facial imitations was examined with a computer vision program (Openface) that performs facial analysis on recorded videos. Results There was no significant difference between the groups in terms of facial emotion recognition accuracy. In group comparisons of voluntary facial mimicry, children with ADHD showed a significantly higher imitation intensity after emotional expressions of sadness, surprise and fear. There was no difference between the groups after the emotions of happiness, anger and disgust. Conclusion This non-obtrusive, noninvasive, and cost-effective method allowed us to measure the quantitative differences in facial mimicry between children with ADHD and typically developing children. Our results contribute new information to the literature by indicating which emotions can be used in the evaluation of social communication skills, as well as intervention targets for these skills, in children with ADHD.

An assessment of dynamic facial emotion recognition and theory of mind in children with ADHD: An eye-tracking study.

Deficits in social cognition in attention deficit hyperactivity disorder (ADHD) have been associated with difficulties in functioning. Since recognizing emotional facial expressions is essential for developing the perceptual components of the theory of mind (ToM), it is important to assess this relationship in children with ADHD. This study therefore compared the recognition of emotional stimuli and gaze patterns between children with ADHD and healthy children using eye-tracking with dynamic facial images. It also examined the relationship between facial emotion recognition accuracy, gaze patterns, ToM scores, and ADHD symptoms. Children with ADHD aged 8-13 (n = 47) and a control group (n = 38) completed a facial emotion recognition test, ToM tests, and the Conners' Parent Rating Scale. Participants' gaze patterns in response to dynamic facial emotion expressions were recorded using eye-tracking technology. Children with ADHD exhibited significantly lower accuracy in the recognition of the facial expressions of disgust and anger. The percentage fixation in the eye region was also significantly lower for happy, angry, sad, disgusted, and neutral emotions in the children with ADHD compared to the control group. No relationship was determined between the percentage of fixations on facial areas of interests and ADHD symptoms or ToM tests. This study provides evidence that children with ADHD experience deficits in visual attention to emotional cues. In addition, it suggests that facial emotion recognition deficits in children with ADHD represent a separate domain of social cognition that develops independently of ToM skills and core symptoms. Understanding and treating the social difficulties of individuals with ADHD may help improve their social functioning.

Pseudo-labeling of transfer learning convolutional neural network data for human facial emotion recognition

The relevance of solving the problem of facial emotion recognition on human images in the creation of modern intelligent systems of computer vision and human-machine interaction, online learning and emotional marketing, health care and forensics, machine graphics and game intelligence is shown. Successful examples of technological solutions to the problem of facial emotion recognition using transfer learning of deep convolutional neural networks are shown. But the use of such popular datasets as DISFA, CelebA, AffectNet, for deep learning of convolutional neuralnetworks does not give good results in terms of the accuracy of emotion recognition, because almost all training sets have fundamental flaws related to errors in their creation, such as the lack of data of a certain class, imbalance of classes, subjectivity and ambiguity of labeling, insufficient amount of data for deep learning, etc. It is proposed to overcome the noted shortcomings of popular datasets for emotion recognition by adding to the training sample additional pseudo-labeled images with human emotions, on which recognition occurs with high accuracy. The aim of the research is to increase the accuracy of facial emotion recognitionon the image of a human by developing a pseudo-labeling method for transfer learning of a deep neural network. To achieve the aim, the following tasks were solved: a convolutional neural network model, previously trained on the ImageNet set using the transfer learning method, was adjusted on the RAF-DB data set to solve emotion recognition tasks; a pseudo-labeling method of the RAF−DB set data was developed for semi-supervised learning of a convolutional neural network model for the task of facial emotion recognition; the accuracy of facial emotion recognition was analyzed based on the developed convolutional neural network model and the method of pseudo-labeling of RAF-DB set data for its correction. It is shown that the use of the developed method of pseudo-labeling data and transfer learning of the MobileNet V1 convolutional neural network model allowed to increase the accuracy of facial emotion recognitionon the images of the RAF-DB dataset by 2 percent (from 76 to 78%) according to the F1 estimate. Atthe same time, taking into account the significant imbalance of the classes, for the 7 main emotions in the trainingset, we have a significant increase in the accuracy of recognizing a few representatives of such emotions as surprise (from 71 to 77%), fearful(from 64 to 69%), sad (from 72 to 76%), angrywith (from 64 to 74%), neutral(from 66 to 71%). The accuracy of recognizing the emotion of happy, which is the most common, decreased (from 91 to 86 %) Thus, it can be concluded that the use of the developed pseudo-labeling method gives good results in overcoming such shortcomings of datasets for deep learning of convolutional neural networks such as lack of data of a certain type, imbalance of classes, insufficient amount of data for deep learning, etc.

Facial expression recognition based on adaptation of the classifier to videos of the user

In this paper, an approach that can significantly increase the accuracy of facial emotion recognition by adapting the model to the emotions of a particular user (e.g., smartphone owner) is considered. At the first stage, a neural network model, which was previously trained to recognize facial expressions in static photos, is used to extract visual features of faces in each frame. Next, the face features of video frames are aggregated into a single descriptor for a short video fragment. After that a neural network classifier is trained. At the second stage, it is proposed that adaptation (fine-tuning) to this classifier should be performed using a small set of video data with the facial expressions of a particular user. After emotion classification, the user can adjust the predicted emotions to further improve the accuracy of a personal model. As part of an experimental study for the RAVDESS dataset, it has been shown that the approach with model adaptation to a specific user can significantly (up to 20 – 50 %) improve the accuracy of facial expression recognition in the video.

Four attachment-based categories of emotion regulation in adolescent psychic troubles.

Emotion regulation is altered in many psychiatric disorders in adolescence, but the understanding of mechanisms that underlie this alteration is still poor. The PERCEPT study explores alexithymia, empathy, facial emotion recognition (FER) and defence mechanisms in a sample of adolescents in psychiatric care (n = 61, 74% of girls, mean age = 15.03 y.o.), in relation with participants' attachment styles. Results revealed correlations between attachment dimensions and all of the emotion regulation variables, suggesting that attachment modalities have functional links with emotional regulation at its different levels: FER accuracy was inversely correlated with avoidant attachment, while affective empathy, difficulty in identifying feelings (alexithymia) and immature as well as neurotic defence mechanisms were positively correlated with anxious attachment. Moreover, attachment categories delineated distinct emotional perception profiles. In particular, preoccupied attachment included adolescents with the highest levels of facial emotion perception (sensitivity and accuracy) and of affective empathy, whereas detached attachment included adolescents with the lowest levels of these variables. Neurotic defence mechanisms and difficulty to identify feelings were correlated with preoccupied attachment; immature defence mechanisms and difficulty to describe feelings to others characterized fearful attachment. These results suggest that attachment categories underlie emotion regulation processes in psychiatric disorders in adolescence. Theoretical and clinical implications are discussed.

A retrospective case-control study of facial emotion recognition in male veterans with chronic schizophrenia and its correlation with interpersonal communication

ObjectiveTo understand the facial emotion recognition of male veterans with chronic schizophrenia and the relationship between facial emotion recognition and interpersonal communication to provide a reference for designing social skills training programmes.MethodFifty-six eligible male patients with chronic schizophrenia who were admitted to our hospital from October 2020 to April 2021 were selected, and 24 healthy people were selected as controls. Facial emotion recognition, social communication skills and self-perceived interpersonal disturbance were assessed using a facial emotion recognition stimulus manual, the Social Skills Checklist (SSC) and the Interpersonal Relationship Integrative Diagnostic Scale (IRIDS). Disease status was assessed using the Positive and Negative Syndrome Scale.ResultsBoth the control group and the patient group had the highest recognition accuracy for neutral faces. The recognition rate for neutral expression was higher in the control group than in the patient group (p = 0.008). The rate of neutral expressions identified as happiness was higher in the patient group than in the control group (p = 0.001). The identification of anger as happiness was higher in the control group than in the patient group (p = 0.026), and the pattern of misidentification was similar between the control group and the patient group. The accuracy of facial emotion recognition was negatively associated with the age of onset (p < 0.05). The recognition accuracy for happiness was negatively associated with negative symptoms, general pathological symptoms and total scale scores (p < 0.05). The total score for expression recognition was negatively associated with the negative symptom subscale scores (p < 0.05), and there was no correlation between expression recognition and positive symptoms (p > 0.05). The recognition accuracy for happiness was negatively correlated with the IRIDS conversation factor (p < 0.05). The recognition accuracy for happiness and anger and the total scores for facial emotion recognition were negatively correlated with the SSC subscale score and the total score (p < 0.05 and p < 0.01, respectively). The main influencing factors on facial emotion recognition were the SSC total score (p < 0.001) and the positive factor score (p = 0.039).ConclusionVeterans with chronic schizophrenia have facial emotion recognition impairments affected by negative symptoms. There is a correlation between facial emotion recognition and interpersonal communication.Highlights1. There are extensive facial expression recognition disorders in schizophrenia.2. The pattern of misidentification was similar in both the control group and the patient group, with the tendency for happiness to be identified as a neutral emotion, anger as happiness, and fear as neutral emotion and anger.3. Based on the comprehensive assessment of social cognitive impairment in patients with schizophrenia, prospective studies of standardised interventions are designed to provide support for clinical practice.

The relevance of social anxiety for understanding social functioning and facial emotion recognition in individuals at clinical high-risk for psychosis.

Individuals at clinical high risk (CHR) for psychosis often experience poor social functioning and impaired facial emotion recognition (FER); however, the impact of frequently comorbid symptoms upon these processes is underexplored. In particular, social anxiety is characteristic of this population and also related to poor social functioning and FER biases, such as misinterpreting neutral faces as negative or threatening; however, little is known about how social anxiety relates to these processes in CHR individuals. The present study examined the overlap of social anxiety, social functioning, and FER accuracy and bias. Participants (CHR N=62, healthy controls N=52) completed the self-report Social Interaction Anxiety Scale (SIAS), Penn Emotion Recognition-40 (ER-40) behavioural task, and interviewer-rated Global Functioning Scale-Social (GFS-S). The ER-40 was used to assess both FER accuracy (e.g., overall number of correct responses) and bias (e.g., mislabelling neutral faces as angry). Consistent with previous research, relative to controls, CHR participants had more social anxiety (d=-1.07), poorer social functioning (d=-1.62), and performed more poorly on the FER task (e.g., d=-.37). Within CHR participants, social anxiety was related to an anger detection bias (r=.28), above and beyond positive symptom severity, which in turn was related to FER accuracy (r=.26) and social functioning (r=-.28). These findings suggest that ongoing work examining social processes within CHR individuals needs to account for social anxiety and that social anxiety may be a useful preventive intervention target.

Facial emotion recognition in adolescent depression: The role of childhood traumas, emotion regulation difficulties, alexithymia and empathy.

Facial emotion recognition (FER) is crucial for effective social competency, and problems in this skill are linked depression during adolescence. In this study, we aimed to find the rates of FER accuracy for negative (fearful, sad, angry, disgusted), positive (happy, surprised), and neutral emotions, and the possible predictors of FER skill for most confusing emotions. A total of 67 drug-naive adolescents with depression (11 boys, 56 girls; 11-17 years) were recruited for the study. The facial emotion recognition test, childhood trauma questionnaire and basic empathy, difficulty of emotion regulation, and Toronto alexithymia scales were used. The analysis demonstrated that adolescents have more difficulties in recognizing negative emotions when compared the positive ones. The most confusing emotion is fear (39.8% of fear was recognized as surprise). Boys have lower fear recognition skill than girls and higher childhood emotional abuse, physical abuse, emotional neglect, and difficulty in describing feelings to predict lower fear recognition skill. For sadness recognition skill, emotional neglect, difficulty in describing feelings, and depression severity were the negative predictors. Emotional empathy has a positive effect on disgust recognition skill. Our findings demonstrated that impairment of FER skill for negative emotions is associated with childhood traumas, emotion regulation difficulties, alexithymia, and empathy symptoms in adolescent depression.

Multimodal Human Facial Emotion Recognition Using DenseNet-161 and Image Feature Stabilization Algorithm

Human Facial Emotion Recognition (FER) is the technology to predict listener’s emotion of static images and videos to uncover data on one's enthusiastic states like happy, sad, frustration, anxiety, surprise, hate and neutral states. It's a part of the affective computing technology, which may be a collaborative area of research on listener’s emotions. The ability of a computer to recognise and understand and frequently explain human emotions and emotional states which relies on computing (AI) technology. The most difficult part of the FER is to plan the various looks with the separate enthusiastic states. The standard steps of Facial Emotion Recognition are i) Face RoI identification ii) Feature Extraction and iii) Emotion Recognition. Convolutional Neural Network models are the most commonly utilised to detect listener emotions. In this paper, an Image Feature Stabilization Algorithm (IFSA) is proposed to improve the efficiency of facial emotion recognition by implementing Deep Convolutional Neural Network (DCNN) model using the Transfer Learning (TL) technique. The architecture entails employing a FER-compatible pre-trained Densenet-161 based DCNN model and then fine-tuning the model for face emotion data. Initially, the dense layer(s) is/are trained, followed by the fine-tuning of each of the pre-trained DCNN blocks, resulting in an improvement in FER accuracy, particularly for difficult front face views like partial view. Experiments when performed on the CK+ dataset by employing a 10-fold cross validation method using various pre-train models like VGG 16, VGG19, ResNet-18, 34, 50, 152, Inception-V3 and DenseNet-161 showed accuracy of 85.9%, 94.6%, 91%, 91%, 91.1%, 95.1%, 97.1% and 98.7% respectively. Thus, Facial Emotion Recognition performed using the finetuned DenseNet-161 architecture demonstrated exceptionally improved accuracy compared to other pretrained models, along with the proposed image feature stabilization algorithm. The proposed architecture using Densenet-161 showed improved accuracy of 98.78% and 97.52% in other challenging FER datasets like KDEF and JAFEE.

An integrated approach for mental health assessment using emotion analysis and scales

AbstractDepression is a prominent cause of mental illness, which could primarily increase early death. It is possible that this is the root of suicidal ideation, and it causes severe impairment in daily life. By detecting human face traits, artificial intelligence (AI) has cleared the road for predicting human emotions. This predictive technique will be used to conduct a preliminary assessment of depression. Prediction is accomplished using a mixture of four modules namely Facial Emotion Recognition (FER), Scales Questionnaire, Speech Emotion Recognition (SER), and Doctor Chat. FER2013 dataset is used for the FER module, while for speech‐based recognition, RAVDESS, TESS, SAVEE, and CREMA‐D are collectively used. To improve the accuracy of the FER, the people in the given image will be fed into a Face API created with TensorFlow JS, which will eventually be given to the proposed model that will recognize human faces in the image. For SER, a python library known as Librosa is used for extracting audio features and it will be fed to the proposed model. The scales module of the app has questionnaires that can be answered, and the result can be generated based on the scores obtained using established scales used in modern psychology such as the HAM‐D, YMRS etc., Though deep learning can predict emotions, the user may choose to speak with a real doctor about the issues to clear up any doubts. The application has a Doctor Chat module, which is essentially a chat bot for interacting with a doctor. Using this module, the users can talk, exchange files, and have their questions answered. The accuracy of FER is 91% whereas for SER, it is 82% on the test sets. The proposed approach produces the highest accuracy for the benchmark dataset. These four modules will work together to produce a homogenous depression report.

Association between fractional amplitude of low-frequency fluctuation (fALFF) and facial emotion recognition ability in first-episode schizophrenia patients: a fMRI study

It was still unclear that the correlation between the resting-state intrinsic activity in brain regions and facial emotion recognition (FER) ability in patients with first-episode schizophrenia (FSZ). Our aim was to analyse the correlation between the fractional amplitude of low-frequency fluctuation (fALFF) and FER ability in FSZ patients. A total of 28 patients with FSZ and 33 healthy controls (HCs) completed visual search tasks for FER ability. Regions of interest (ROIs) related to facial emotion were obtained from a previous meta-analysis. Pearson correlation analysis was performed to understand the correlation between fALFF and FER ability. Our results indicated that the patients performed worse than the HCs in the accuracy performances of happy FER and fearful FER. The previous meta-analysis results showed that the brain regions related to FER included the bilateral amygdala (AMY)/hippocampus (HIP), right fusiform gyrus (FFG), and right supplementary motor area (SMA). Partial correlation analysis showed that the fALFF of the right FFG was associated with high-load fearful FER accuracy (r = − 0.60, p = 0.004). Our study indicated that FER ability is correlated with resting-state intrinsic activity in brain regions related to facial emotion, which may provide a reference for the study of FER deficiency in schizophrenia.

Social cognition training improves recognition of distinct facial emotions and decreases misattribution errors in healthy individuals.

Facial emotion recognition is a key component of social cognition. Impaired facial emotion recognition is tied to poor psychological wellbeing and deficient social functioning. While previous research has demonstrated the potential for social cognition training to improve overall facial emotion recognition, questions remain regarding what aspects of emotion recognition improve. We report results from a randomized controlled trial that evaluates whether computerized social cognition training can improve recognition of distinct facial emotions in healthy participants. This investigation was designed to better understand the therapeutic potential of social cognition training for individuals with neuropsychiatric disorders. Fifty-five healthy adult participants were randomly assigned to an internet-based intervention during which they either completed social cognition training (SCT) or played control computer games (CON) for 10.5 h over 2-3 weeks. Facial emotion recognition was measured with the Penn ER-40, which was conducted before and after training. The following variables were collected and analyzed: facial emotion recognition accuracy for each emotion (i.e., anger, fear, happy, neutral (no emotional expression), and sad), reaction times for each emotion, and response error types (i.e., frequency of an emotion being chosen incorrectly, frequency of an emotion being missed, and frequency of an emotion being confused for another particular emotion). ANOVAs and t-tests were used to elucidate intervention effects both within and between groups. Results showed that the SCT group improved their accuracy for angry and neutral faces. They also improved their reaction times for neutral, fearful, and sad faces. Compared to the CON group, the SCT group had significantly faster reaction times to neutral faces after training. Lastly, the SCT group decreased their tendency to confuse angry faces for no emotional expression and to confuse no emotional expression for sad faces. In contrast, the CON group did not significantly improve their accuracy or reaction times on any emotional expression, and they did not improve their response error types. We conclude that social cognition training can improve recognition of distinct emotions in healthy participants and decrease response error patterns, suggesting it has the potential to improve impaired emotion recognition and social functioning in individuals with facial emotion recognition deficits.

Research on Classroom Emotion Recognition Algorithm Based on Visual Emotion Classification.

In this paper, we construct a classroom emotion recognition algorithm by classifying visual emotions for improving the quality of classroom teaching. We assign weights to the training images through an attention mechanism network and then add a designed loss function so that it can focus on the feature parts of face images that are not obscured and can characterize the target emotion, thus improving the accuracy of facial emotion recognition under obscuration. Analyze the salient expression features of classroom students and establish a classification criteria and criteria library. The videos of classroom students' facial expressions are collected, a multi-task convolutional neural network (MTCNN) is used for face detection and image segmentation, and the ones with better feature morphology are selected to build a standard database. A visual motion analysis method with the fusion of overall and local features of the image is proposed. To validate the effectiveness of the designed MTCNN model, two mainstream classification networks, VGG16 and ResNet18, were tested and compared with MTCNN by training on RAF-DB, masked dataset, and the classroom dataset constructed in this paper, and the final accuracy after training was 78.26% and 75.03% for ResNet18 and VGG16, respectively. The results show that the MTCNN proposed in this paper has a better recognition effect. The test results of the loss function also show that it can effectively improve the recognition accuracy, and the MTCNN model has an accuracy of 93.53% for recognizing students' facial emotions. Finally, the dataset is extended with the training method of expression features, and the experimental study shows that the method performs well and can carry out recognition effectively.

A novel enhanced convolution neural network with extreme learning machine: facial emotional recognition in psychology practices

Facial emotional recognition is one of the essential tools used by recognition psychology to diagnose patients. Face and facial emotional recognition are areas where machine learning is excelling. Facial Emotion Recognition in an unconstrained environment is an open challenge for digital image processing due to different environments, such as lighting conditions, pose variation, yaw motion, and occlusions. Deep learning approaches have shown significant improvements in image recognition. However, accuracy and time still need improvements. This research aims to improve facial emotion recognition accuracy during the training session and reduce processing time using a modified Convolution Neural Network Enhanced with Extreme Learning Machine (CNNEELM). The system entails (CNNEELM) improving the accuracy in image registration during the training session. Furthermore, the system recognizes six facial emotions happy, sad, disgust, fear, surprise, and neutral with the proposed CNNEELM model. The study shows that the overall facial emotion recognition accuracy is improved by 2% than the state of art solutions with a modified Stochastic Gradient Descent (SGD) technique. With the Extreme Learning Machine (ELM) classifier, the processing time is brought down to 65ms from 113ms, which can smoothly classify each frame from a video clip at 20fps. With the pre-trained InceptionV3 model, the proposed CNNEELM model is trained with JAFFE, CK+, and FER2013 expression datasets. The simulation results show significant improvements in accuracy and processing time, making the model suitable for the video analysis process. Besides, the study solves the issue of the large processing time required to process the facial images.

Prediction of Facial Emotion Recognition Ability in Patients With First-Episode Schizophrenia Using Amplitude of Low-Frequency Fluctuation-Based Support Vector Regression Model.

ObjectiveThere were few studies that had attempted to predict facial emotion recognition (FER) ability at the individual level in schizophrenia patients. In this study, we developed a model for the prediction of FER ability in Chinese Han patients with the first-episode schizophrenia (FSZ).Materials and MethodsA total of 28 patients with FSZ and 33 healthy controls (HCs) were recruited. All subjects underwent resting-state fMRI (rs-fMRI). The amplitude of low-frequency fluctuation (ALFF) method was selected to analyze voxel-level spontaneous neuronal activity. The visual search experiments were selected to evaluate the FER, while the support vector regression (SVR) model was selected to develop a model based on individual rs-fMRI brain scan.ResultsGroup difference in FER ability showed statistical significance (P < 0.05). In FSZ patients, increased mALFF value were observed in the limbic lobe and frontal lobe, while decreased mALFF value were observed in the frontal lobe, parietal lobe, and occipital lobe (P < 0.05, AlphaSim correction). SVR analysis showed that abnormal spontaneous activity in multiple brain regions, especially in the right posterior cingulate, right precuneus, and left calcarine could effectively predict fearful FER accuracy (r = 0.64, P = 0.011) in patients.ConclusionOur study provides an evidence that abnormal spontaneous activity in specific brain regions may serve as a predictive biomarker for fearful FER ability in schizophrenia.

Facial emotion recognition difficulties may be specific to skin picking disorder, but could also be related to the presence of alexithymia in trichotillomania

Background and objectivesPrevious research has shown high rates of alexithymia and emotion dysregulation in trichotillomania (TTM) and skin picking disorder (SPD). Unfortunately, there are no data on facial emotion recognition (FER) in TTM and SPD. The present study aimed to compare patients with TTM and SPD and a healthy control group for the severity of alexithymia and rates of FER. MethodsForty patients with SPD, 30 patients with TTM, and 30 healthy controls were enrolled in this study. The Beck Depression Inventory (BDI), Beck Anxiety Inventory (BAI), Clinical Global Impression (CGI), Toronto Alexithymia Scale (TAS-20), and the Facial Emotion Recognition Test were applied to the participants. ResultsPatients with TTM and SPD had less FER accuracy and higher alexithymia scores compared with healthy controls. According to ANCOVA analysis, when anxiety, depression, and alexithymia were fixed as covariates, disgusted facial expressions and total facial emotion recognition were still significantly lower in patients with SPD compared with the control group, but there was no difference between the TTM and control groups and TTM and SPD groups. ConclusionAlexithymia rates were high in patients with TTM and SPD. Interestingly, difficulty in recognizing disgusted facial expressions may be a distinctive sign in SPD. Future neuroimaging studies are needed to support possible FER impairment in patients with TTM/SPD.