Search Results (138)

This study examined whether AI-powered chatbot-based training was associated with changes in university students’ Facial and Vocal Emotional Reaction Time Proportions during simulated job interviews. A one-group pretest–posttest design was conducted with 54 third- and fourth-year students enrolled in a Human Talent Management course at a private Latin American university. This study was implemented in an Applied Neuroscience Laboratory using iMotions-supported facial-expression recognition and vocal analysis technologies. Participants first completed a baseline simulated interview, followed by three chatbot-based training sessions using HR-expert-validated questions, end-of-session scoring, and qualitative feedback. A final simulated interview was then conducted to compare pre- and post-training indicators. Facial emotional reaction time was analyzed through aggregate indicators—positive, negative, neutral, confusion, and sentimentality—and specific facial-expression categories, including joy, surprise, anger, sadness, disgust, fear, and contempt. Vocal emotional reaction time was examined through happiness, sadness, anger, and neutrality. Pre–post differences were assessed using paired-samples t-tests and complementary Wilcoxon signed-rank tests. Positive facial emotional reaction time increased significantly from 3.52% to 14.75%, with a mean increase of 11.23 percentage points, 95% CI [4.79, 17.67]. Facial joy increased significantly from 2.38% to 10.10%, with a mean increase of 7.72 percentage points, 95% CI [3.30, 12.14], while vocal happiness increased significantly from 2.79% to 10.71%, with a mean increase of 7.92 percentage points, 95% CI [3.38, 12.46]. Each of these principal outcomes showed a standardized paired effect of d_z = 0.48, 95% CI [0.19, 0.76], with corresponding Wilcoxon effect-size estimates ranging from r = 0.40 to r = 0.44. Several negative and neutral indicators also decreased after training; however, their mean-based standardized effects were generally smaller and some statistically significant findings were supported primarily by the Wilcoxon signed-rank test. Overall, chatbot-based interview training was associated with changes in algorithmically classified facial and vocal-expression patterns and may provide a complementary tool for structured interview practice in higher education. Full article

Background: The cerebellum has recently been proposed as a key contributor to facial expression (FE) processing, although the neural pathways underlying its involvement remain poorly understood. This study investigated whether cerebellar modulation of FE processing preferentially relies on motor or perceptual circuits. Methods: Sixteen healthy young participants underwent cerebellar cathodal transcranial direct current stimulation (tDCs) and sham stimulation in a randomized crossover design. Cerebellar–motor cortex interactions were assessed using cerebellar brain inhibition (CBI) measured by a paired-pulse transcranial magnetic stimulation protocol, while early perceptual processing was evaluated through P100 and N170 event-related potentials (ERPs). Behavioral performance was assessed using a task involving recognition of neutral, happy, and fearful faces. Results: Cerebellar tDCs selectively modulated the cerebello–thalamo–primary motor cortex pathway, producing a significant reduction in CBI during the passive viewing of emotional (happy and fearful) expressions while exerting no significant effects on occipitotemporal ERP components associated with early face perception or on behavioral measures of emotion recognition, including recognition accuracy and reaction times. Conclusions: These findings indicate that cerebellar contributions to facial emotion processing are predominantly mediated through motor rather than early perceptual pathways. The cerebellum appears to modulate motor network activity during emotional face observation without directly influencing occipitotemporal perceptual mechanisms, although this physiological modulation was not accompanied by measurable changes in behavioral performance. These findings support the hypothesis that cerebellar involvement in social cognition is primarily related to the optimization of emotion-related motor processing. Full article

Emotion recognition remains a challenging task despite substantial progress in machine learning and affective computing. This study examines challenges in emotion recognition through a comparative analysis of two widely used modalities: facial images and speech signals. The analysis was conducted using FER-2013 for facial emotion recognition and the TESS and RAVDESS datasets for speech emotion recognition. A MobileNetV2-based approach was applied to visual data, while speech analysis employed MFCC-based representations and both classical and deep learning models. The study combines quantitative performance evaluation with qualitative analysis of classification behavior, focusing on emotion-specific recognition difficulties and recurring error patterns across modalities. Model performance was assessed using accuracy, precision, recall, F1-score, and confusion matrices. Across the analysed datasets, overall classification accuracy ranged from approximately 73% to 96%, while class-level F1-scores ranged from 0.48 to 0.89 depending on the emotion and modality. Happiness and surprise consistently achieved the highest recognition performance, whereas neutral emotion, fear, and disgust exhibited the lowest class-level F1-scores and generated the highest numbers of misclassifications. The experimental results confirmed that happiness and surprise achieved the highest classification performance across modalities, while neutral emotion, fear, and disgust showed reduced recognition accuracy due to weak expressive cues and overlapping feature representations. These difficulties are associated with weak or ambiguous expressive signals, overlap between emotional categories, and variability in emotional expression. The comparative findings suggest that recognition challenges arise from both modality-specific limitations and the inherent properties of emotional expression. The results highlight the importance of multimodal approaches and more flexible representations for improving emotion recognition systems. Full article

Individuals differ in their susceptibility to emotional contagion, i.e., the automatic tendency to mirror and synchronize another person’s expressions and movements, resulting in shared emotional experiences. The objective of this research was to investigate how susceptibility to emotional contagion connects to automatic facial emotion processing and emotional competences. An affective priming task using happy, angry, neutral, and blank faces was administered to a sample of 104 women with a mean age of 24.72 years (SD = 3.63). They completed self-report measures assessing susceptibility to positive and negative emotional contagion, alexithymia, trait emotional intelligence, affectivity, and depression. Although prime valence-congruent evaluative shifts were found in our sample, there were no correlations of susceptibility to positive and negative emotional contagion with affective priming effects. Susceptibility to positive emotion contagion was negatively correlated with alexithymia and positively with emotional intelligence. However, susceptibility to positive emotion contagion predicted only emotional intelligence (but not alexithymia), when controlling for relevant affect variables. Our findings indicate that emotional contagion susceptibility could be less strongly linked to automatic emotion perception than previously suggested. Moreover, the trait-like tendency to resonate with other people’s positive emotions seems to be linked to enhanced capacities in perceiving, interpreting, and regulating emotions. Full article

Emotional recognition relies on the integration of multiple affective cues. In everyday contexts, however, facial expressions, vocal prosody, and semantic content may convey incongruent emotional information, generating emotional conflict and increasing cognitive demands. The present study examined how multimodal emotional conflict affects emotion recognition during video viewing, focusing on short videos in which a single actor simultaneously conveyed incongruent emotional cues across facial, vocal, and semantic channels. Forty-seven undergraduate students completed a gaze-based response task in which, after each short video, they provided a single judgment of the overall emotion conveyed by the stimulus. The videos depicted either congruent or incongruent combinations of semantic content, facial expressions, and vocal prosody across six basic emotions and a neutral condition. Data were analyzed using repeated-measures ANOVAs and generalized linear mixed-effects models. Accuracy was consistently higher for congruent than incongruent stimuli across all domains, indicating a robust emotional interference effect. Critically, the magnitude of this effect differed by domain. Semantic content showed the largest performance reduction under incongruence, followed by facial expression and vocal prosody. Mixed-effects models confirmed these effects while accounting for participant- and item-level variability and revealed a significant Congruency × Domain interaction. In a gaze-based response task requiring a single overall emotion judgment, emotional conflict disrupted recognition in a domain-specific manner, with semantic information being particularly vulnerable to multimodal interference. Full article

This study examined the interaction between spatial frequencies and emotion processing using tachistoscopic presentations of emotional faces, in a patient with right capsulo–thalamic damage and a matched control group (N = 3). Emotional (happy, angry and sad) and neutral faces were presented in one of two ways: broadband emotional images and hybrid faces, which were created by superimposing emotional Low Spatial Frequencies (LSFs) to the High Spatial Frequencies (HSFs) of the same identity with a neutral expression, resulting in a subliminal presentation of the emotional content. According to LeDoux’s dual-route model, which suggests a cortical–conscious emotional analysis and subcortical–unconscious emotional processing, we expected healthy participants to show different variations in friendliness ratings compared with the case study patient. In particular, we hypothesized that while healthy participants should show friendliness ratings varying consistently with the facial expressions for both unfiltered (conscious) and filtered (unconscious) stimuli, reflecting the efficiency of both routes, the patient should show a selective deficit in the unfiltered condition due to the disruption of the thalamo–cortical connections. The results showed that healthy controls evaluated emotions consistently across both conditions. Notably, there were no significant differences between the case study patient and the control group for hybrid faces, suggesting that the “hidden” LSF successfully activated the intact subcortical route. However, significant differences emerged for unfiltered stimuli: the case study patient was able to distinguish between positive and negative valence, but she failed to discriminate between negative emotions. This finding suggests that the fine-grained differentiation of negative emotions requires an intact cortical analysis, mediated by the internal capsule. Full article

Automated recognition of equine–human interaction behaviors from video represents a significant challenge in computational ethology, with critical applications spanning animal welfare assessment, equine-assisted services evaluation, and safety monitoring in equestrian environments. Existing approaches to animal behavior recognition typically focus on single species in isolation, rely solely on facial expression analysis while ignoring full-body posture, or employ flat classification architectures that fail under the severe class imbalances characteristic of naturalistic behavioral datasets. Furthermore, no prior framework integrates simultaneous analysis of both human and equine body language for cross-species interaction classification. This paper presents a novel hierarchical classification framework integrating multi-modal computer vision features to distinguish behavioral states during horse–human encounters. Our methodology employs three complementary feature extraction pipelines: YOLOv8 for spatial relationship modeling, MediaPipe for human postural analysis, and AP-10K for equine body language interpretation. From 28 annotated interaction videos comprising 50,270 temporal samples across five horse breeds, we extract 35 discriminative features capturing proximity dynamics, body orientation, and species-specific behavioral indicators. To address severe class imbalance (18.3:1 ratio between affiliative and avoidant categories), we implement cost-sensitive gradient boosting with automatic class weight optimization within a two-stage hierarchical architecture. The first stage classifies interactions into three parent categories (affiliative, neutral, avoidant) achieving 73.2% balanced accuracy, while stage two discriminates six fine-grained sub-behaviors achieving 88.5% balanced accuracy (under oracle parent-category routing; cascaded end-to-end performance is 62.9% balanced accuracy due to Stage 1 error propagation, identifying parent classification as the primary bottleneck). Notably, our system achieves 85.0% recall on safety-critical avoidant behaviors despite their representation of only 3.8% of the dataset. Extensive ablation studies demonstrate that equine pose features contribute most critically to classification performance, while comprehensive cross-validation analysis confirms model robustness across diverse interaction contexts. The proposed framework establishes the first systematic multimodal cross-species behavioral assessment pipeline in human–animal interaction research, with direct implications for improving equine welfare monitoring and rider safety protocols. Full article

Background: This study examined age-related differences and interrelationships among psychological symptoms, personality traits, and emotional expression styles in a community sample of 151 participants aged 10–77 years, spanning four age groups: adolescents, young adults, middle-aged adults, and older adults. Methods: Psychological symptoms were assessed using the SCL-90, personality traits using the Big Five Inventory-2 (BFI-2), and emotional expression patterns were derived from facial expression recognition via a convolutional neural network (CNN) model. Kruskal–Wallis H tests were used to examine age-related differences. K-means cluster analysis was applied to identify emotional expression patterns, and logistic regression was used to construct a mental health risk screening model. Results: The young adult group (19–35 years) achieved the highest scores on the depression (M = 1.73) and anxiety (M = 1.61) dimensions, indicating a higher level of psychological distress during this life stage. Personality traits showed a significant developmental trajectory: neuroticism decreased with age (H(3) = 17.09, p < 0.001, η² = 0.11), declining from 2.69 in the young adult group to 2.17 in the older adult group; conscientiousness increased with age (H(3) = 37.39, p < 0.001, η² = 0.24), representing the most substantial age-related effect. K-means clustering identified three distinct emotional expression patterns: Cluster 1 was characterised by happiness, Cluster 2 by anger, disgust, and fear, and Cluster 3 by neutrality, sadness, and surprise. Cluster 2 exhibited the highest scores on neuroticism, anxiety, depression, and mood swings, and scored significantly higher than the other two clusters on interpersonal sensitivity, depression, anxiety, and hostility (p < 0.05). Mental health risk screening indicated that 26.5% of participants were classified as high-risk. Logistic regression analysis (AUC = 0.742) showed that neuroticism was the strongest predictor of elevated mental health risk (OR = 4.58), while extraversion (OR = 0.41) and conscientiousness (OR = 0.57) were significant protective factors. Conclusions: These findings provide exploratory evidence regarding age-related patterns of psychological symptoms and personality traits in a convenience sample and offer preliminary support for personality-based mental health risk screening. Notably, the SCL-90 was employed as a screening tool rather than for clinical diagnosis. Given the unequal age group sizes, particularly the small young adult subgroup, generalisability across the lifespan should not be assumed. Full article

Emotional and workload-related states unfold dynamically during immersive virtual reality (VR) experiences, yet reliable physiological modeling in such environments remains challenging. We investigated whether multi-channel facial electromyography (fEMG), combined with spatio-temporal deep learning, can (i) accurately classify calibrated facial expressions across participants and (ii) transfer to spontaneous, task-elicited behavior in immersive VR. Twelve adults completed a calibration phase involving four intentional expressions (smile, frown, raised eyebrow, neutral), followed by VR scenes designed to elicit emotional, cognitive, physical, and dual task demands. After participant-level physiological normalization, a single shared Convolutional Neural Network–Temporal Convolutional Network (CNN–TCN) model was trained and evaluated using leave-one-participant-out (LOPO) validation. The model achieved strong cross-participant performance (Macro-F1 = 0.88 ± 0.13; ROC-AUC = 0.95 ± 0.06). When applied to unlabeled spontaneous VR task-elicited fEMG recordings, the trained model generated continuous expression classes. Derived static and temporal expression features showed scene-dependent modulation and False Discovery Rate (FDR)-surviving associations, primarily with perceived physical demand (NASA-TLX). The observed muscle activation patterns were physiologically plausible and aligned with Facial Action Coding System (FACS)-based interpretations of underlying muscle activity. These findings demonstrate that end-to-end spatio-temporal modeling of raw fEMG enables facial expression sensing in immersive VR using a single shared model following physiological normalization. The proposed framework bridges calibrated expression learning and spontaneous task-elicited behavior, supporting privacy-preserving, continuous and physiologically grounded monitoring in human-centered VR applications. Full article

Facial expression recognition (FER) in naturalistic settings is constrained by label ambiguity and variability in stimulus–response alignment. Adopting a data-centric perspective, this study examined whether emotional intelligence (EI)-stratified training data influence FER performance by treating EI as a qualitative factor associated with affective data consistency. Naturally elicited facial expressions were collected in a controlled emotion induction experiment with subjective arousal and valence ratings. Using response-driven labeling, neutral ratings were retained as indicators of ambiguity. Participants were grouped into High and Low EI based on the alignment between subjective evaluations and outputs from a pretrained affect estimator. Identical binary classifiers for arousal and valence recognition were trained while varying only the training data composition and evaluated across baseline, unambiguous, and ambiguous test sets using independent training repetitions with repetition-level statistical aggregation. EI-stratified training was associated with statistically detectable, context-dependent performance differences: group effects were observed primarily under baseline conditions and, to a lesser extent, under ambiguous conditions, whereas no reliable differences emerged under unambiguous conditions. Pooled discrimination differences were modest, but item-level analyses identified significant differences in classification correctness in specific task–condition combinations. Comparable patterns were observed across alternative backbone architectures. These findings indicate that FER performance in naturalistic contexts is influenced not only by model architecture but also by the statistical structure and internal coherence of the training data, supporting EI-informed data selection in ambiguity-prone scenarios. Full article

Objectives: The purpose of the study was to test the hypothesis that high anxiety is associated with biased processing of threat-related stimuli and that anxious individuals may be particularly sensitive to facial expressions of fear or anger. In addition, these effects may result from a specific pattern occurring in the early stages of visual information processing. Methods: Event-Related Potentials (ERPs) were recorded in response to task-irrelevant pictures of faces presented in either an upright or inverted position in two groups differing in trait anxiety, as assessed by scores on the Spielberger Trait Anxiety Inventory (STAI). Behavioural responses and ERP activity were also recorded in response to simple neutral visual stimuli presented during exposure to the facial stimuli, which served as probe-targets. Results: A typical Face Inversion Effect was observed, characterised by longer latencies and greater amplitudes of the early P1 and N170 ERP components. Differences between low- and high-anxious individuals emerged at parieto-occipital sites within the time window of the early P1 component. The later stage of face processing, indexed by the N170 component, was not affected by the level of trait anxiety. Conclusions: The results of this experiment indicate that anxiety level modulates the initial stages of information processing, as reflected in the P1 component. This may be associated with anxiety-related differences in the involuntary processing of face detection of emotional expression. Consequently, a greater attentional engagement appears to occur in highly anxious individuals, leading to delayed behavioural responses to concurrently presented neutral stimuli. Full article

Background/Objectives: Facial expressions are central to nonverbal communication and social cognition, and their recognition is shaped not only by facial features but also by contextual cues and prior experience. In high-threat contexts, rapid and accurate decoding of others’ emotions is adaptively advantageous. Grounded in neurocognitive models of face processing and vigilance, we tested whether brief perceptual adaptation to emotionally salient scenes, real-world disaster exposure, and pre-traumatic stress reactions enhance facial-emotion categorization. Methods: Fifty healthy adults reported prior direct exposure to catastrophic events (present/absent) and completed the Pre-Traumatic Stress Reactions Checklist (Pre-Cl; low/high). In a computerized task, participants viewed a single adaptor image for 5 s—negative (disaster), positive (pleasant environment), or neutral (phase-scrambled)—and then categorized a target face as emotional (fearful, angry, happy) or neutral as quickly and accurately as possible. Performance was compared across adaptation conditions and target emotions and examined as a function of disaster exposure and Pre-Cl. Results: Emotional adaptation (negative or positive) yielded better performance than neutral adaptation. Higher-order interactions among adaptation condition, target emotion, disaster exposure, and Pre-Cl indicated that the magnitude of facilitation varied across specific facial emotions and was modulated by both experiential (exposed vs. non-exposed) and dispositional (low vs. high Pre-Cl) factors. These effects support a combined influence of short-term contextual tuning and longer-term experience on facial-emotion categorization. Conclusions: Brief exposure to emotionally salient scenes facilitates subsequent categorization of facial emotions relative to neutral baselines, and this benefit is differentially shaped by prior disaster exposure and pre-traumatic stress. The findings provide behavioral evidence that short-term perceptual adaptation and longer-term experiential predispositions jointly modulate a fundamental communicative behavior, consistent with neurocognitive accounts in which context-sensitive visual pathways and salience systems dynamically adjust to support adaptive responding under threat. Full article

With the rapid development of artificial intelligence technology, facial expression recognition (FER) has gained increasingly widespread applications in digital human generation, humanoid robotics, mental health, and human–computer dialogue. Typical FER algorithms based on machine learning have been widely studied over the past few decades, which motivated our survey. In this study, we have surveyed the state of the art in FER across two categories: traditional machine learning-based (ML-based) and deep learning-based (DL-based) approaches. Each category is analyzed based on six subcategories. Then, twelve methods, including four ML-based models and eight DL-based models, are compared to evaluate FER performance across four datasets. The experimental results show that in validation sets, the average accuracy of HOG-SVM is 50.12%, which is the best performance for the four ML-based methods; in contrast, Poster has an average accuracy of 75.98%, which is the best result obtained among the eight DL-based methods. The most difficult expression to recognize is contempt, with recognition accuracies of 10.00% and 40.06% for ML-based and DL-based methods, respectively. The accuracy of the ML-based method for identifying neutral expression is the highest at 35.25%; the DL-based method has the highest accuracy in identifying surprise at 69.56%. From the theoretical analysis and comparative experimental results of existing methods, we can see that FER faces challenges, including inaccurate recognition in complex environments and unbalanced data categories, highlighting several future research directions, especially those involving the latest applications of digital humans and large language models. Full article

Background: Emotional prosody refers to the variations in pitch, pause, melody, rhythm, and stress of pronunciation conveying emotional meaning during speech. Although several studies demonstrated that the cerebellum is involved in the network subserving recognition of emotional facial expressions, there is only preliminary evidence suggesting its possible contribution to recognising emotional prosody by modulating the activity of cerebello-prefrontal circuits. The present study aims to further explore the role of the left and right cerebellum in the recognition of emotional prosody in a sample of healthy individuals who were required to identify emotions (happiness, anger, sadness, surprise, disgust, and neutral) from vocal stimuli selected from a validated database (EMOVO corpus). Methods: Anodal transcranial Direct Current Stimulation (tDCS) was used in offline mode to modulate cerebellar activity before the emotional prosody recognition task, and functional near-infrared spectroscopy (fNIRS) was used to monitor stimulation-related changes in oxy- and deoxy- haemoglobin (O2HB and HHB) in prefrontal areas (PFC). Results: Right cerebellar stimulation reduced reaction times in the recognition of all emotions (except neutral and disgust) as compared to both the sham and left cerebellar stimulation, while accuracy was not affected by the stimulation. Haemodynamic data revealed that right cerebellar stimulation reduced O2HB and increased HHB in the PFC bilaterally relative to the other stimulation conditions. Conclusions: These findings are consistent with the involvement of the right cerebellum in modulating emotional processing and in regulating cerebello-prefrontal circuits. Full article

The present study examined associations between presence/absence of a mask and facial emotion recognition, and how interoceptive awareness, i.e., the perception of internal bodily sensations, may influence associations between them. Eighty-two university students participated in an online behavioral experiment. Participants were required to evaluate categories of emotions as well as valence and arousal levels of facial stimuli that were either neutral or expressed one of Paul Ekman’s basic emotions, i.e., anger, disgust, fear, happiness, sadness, and surprise. Participants also completed a psychological scale on interoceptive awareness. Results showed that accuracy of categorization was significantly lower and levels of valence and arousal were significantly closer to neutral in masked than in unmasked faces for multiple emotions. In addition, individuals who showed higher, as compared to lower, emotional awareness reported significantly higher levels of valence for masked stimuli that expressed surprise. These results suggest that wearing a mask can impair accuracy of facial emotion recognition and sensitivity to emotions, whereas awareness of the association between interoception and emotion might mitigate impairments of sensitivity to emotions in masked faces. Full article