Search Results (119)

There are still challenges in speech emotion recognition, as the representation capability of single-modal information is limited, there are difficulties in capturing continuous emotional transitions in discrete emotion annotations, and the issues of modal structural differences and cross-sample alignment in multimodal fusion methods persist. To address these, this study undertakes work from both data and model perspectives. For data, a Chinese multimodal database STEM-E²VA was constructed, synchronously collecting four modalities of data: articulatory kinematics, acoustics, glottal signals, and videos. This covers seven discrete emotion categories and employs PAD continuous annotation. By integrating discrete and continuous dimensional annotations, it better represents the distinction between strong and weak emotions under the same discrete emotion label. Concurrently, to process the biases in PAD annotations, we employed the SCL-90 psychological questionnaire to analyze annotators’ cognitive and emotional perceptions, thereby ensuring data reliability. For model, this paper proposes a multimodal supervised contrastive fusion network incorporating PAD perception. It employs a PAD-enhanced hybrid contrastive loss function to optimize intra-model and inter-modal feature alignment. Utilizing a cross-attention mechanism combined with a GRU–Transformer network for temporal feature extraction, it achieves deep fusion of multimodal information, reducing inter-modal discrepancies and cross-class confusion. Experiments demonstrate that the proposed method achieves 85.47% accuracy in discrete sentiment recognition on STEM-E²VA, with a substantial reduction in RMSE for PAD dimension prediction. It also exhibits excellent generalization capability on IEMOCAP, providing a novel framework for integrating discrete and continuous sentiment representations. Full article

Driven by the development of next-generation wireless networks and the widespread adoption of sensing, communication is shifting from traditional bit-level transmission to intelligent, rich interactions within our digital social system. However, existing speech semantic communication frameworks predominantly focus on textual accuracy, neglecting the critical affective information (e.g., tone and emotion) that is essential for natural human-centric interactions in the real world. To address this limitation, we propose the Heterogeneous Affective Speech Semantic Communication (HASCom) framework, designed for the robust transmission of highly expressive speech over complex wireless channels. Specifically, we design a heterogeneous dual-stream transmission architecture that decouples discrete phoneme-level linguistic content from continuous emotional embeddings. For discrete semantic information, we use reliable digital coding protected by Low-Density Parity-Check (LDPC) to guarantee strict recoverability. Conversely, for emotional features, we employ Deep Joint Source-Channel Coding (JSCC) analog transmission to prevent irreversible quantization errors and the cliff effect. Additionally, we develop a prior-guided diffusion reconstruction module at the receiving end. This module leverages a structural prior network to align the decoded semantics, which then steers the reverse diffusion process conditioned on the recovered affective features. Extensive experiments under both AWGN and Rayleigh fading channels demonstrate that HASCom significantly outperforms state-of-the-art baselines. Specifically, it achieves superior objective semantic similarity and subjective Mean Opinion Score (MOS) at low Signal-to-Noise Ratios (SNRs), while the JSCC transmission modules maintain an ultra-low inference latency of less than 0.1 ms, validating its high efficiency and robustness for practical deployments. Full article

Reliable evaluation of emotional expression in large language model (LLM) outputs remains methodologically under-specified, particularly for long-form generation where label-only correctness provides limited evidence of affective reliability. A claim-conditioned framework is introduced for cross-model comparison under matched elicitation conditions, with TEAS (Text Emotion Adherence Score) as its core continuous metric. Defined in a shared prototype space induced by a frozen reference encoder, TEAS combines affective separability with entropy-aware uncertainty, enabling reliability assessment beyond discrete agreement within a fixed evaluator. Evaluation is conducted on a controlled synthetic corpus under a ground-truth-free, claim-conditioned protocol across four widely used LLM families (Gemini, GPT, Grok, and Mistral). In addition to overall comparative ordering, auxiliary diagnostic measures are reported to localize failure modes and support interpretation of model behavior, together with Holm-corrected pairwise comparisons, sequence-level drift analysis, and local hyperparameter sensitivity analysis. Empirical results show stable endpoint separation, aggregation-sensitive differences among close models, measurable sequence-level degradation, and stable relative orderings under tested local parameter variations. Overall, the study provides an interpretable and statistically grounded protocol for assessing emotion-expression reliability in LLM-generated text within a fixed reference space rather than as a human gold measure of emotional truth. Full article

Background/Objectives: Healthcare professionals operate in complex and demanding environments characterized by high workloads, emotional strain, and organizational pressures that can undermine well-being. According to Self-Determination Theory, the fulfillment of core psychological needs (autonomy, competence, and relatedness) leads to increased job satisfaction, a key indicator of occupational well-being. Additionally, leadership plays a central role in shaping needs-fulfilling environments. Drawing on Leader–Member Exchange Theory (LMX), which emphasizes that high-quality leader-follower relationships foster greater discretion, provide learning opportunities, and build constructive team interactions, this study aimed to examine whether supportive leadership is associated with job satisfaction through the mediation of autonomy, team task cohesion, and perceived training opportunities. Methods: Data were collected from a local health authority in Northern Italy through an anonymous online survey, completed by 697 healthcare professionals, including 546 non-medical healthcare staff (primarily nurses) and 151 physicians. Structural equation modeling with a robust maximum likelihood estimator was employed to test the mediation model, including professional role as a covariate. Results: Higher LMX was positively and directly associated with job satisfaction, through the partial mediation of autonomy, team cohesion, and training opportunities, all positively associated with satisfaction. Team task cohesion showed the strongest associations with both LMX and satisfaction. Physicians reported slightly higher levels of autonomy, training opportunities, and job satisfaction than non-medical professionals. Conclusions: The findings suggest that supportive leadership contributes to healthcare professionals’ job satisfaction both directly and indirectly by contributing to core needs fulfillment. Interventions that strengthen relational quality, promote team cohesion, and enhance professional development may help sustain well-being and adaptive functioning in high-demand healthcare environments. Full article

Organizations communicate across many channels, yet official websites remain a controlled, authoritative space where firms articulate identity and strategy. This study examines how Croatia’s top enterprises (n = 100) describe themselves on their websites and which emotional tones they use to signal strategic intent. Our goal is to identify recurring strategic signifiers and map distinct sentiment profiles in corporate narratives. We compiled company descriptions from official sites; texts were originally in Croatian and machine-translated into English, and all analysis was conducted on the English corpus. Using lexicon-based sentiment methods (AFINN, Bing, NRC), we quantified polarity and discrete emotions, aggregated scores at the firm level, and applied k-means clustering to normalized emotion vectors. Results show a consistent emphasis on mission–vision–values language and a dominance of positive emotions—especially trust and anticipation. We interpret, based on cluster exemplars, that higher trust/anticipation tones can function as soft governance cues, while transparency about negatives characterizes an issue-addressing regime without eroding overall positivity. Cluster analysis reveals three stable profiles: optimistic consumer-oriented narratives, transparent issue-addressing messaging, and low-affect technical descriptions. We conclude that sentiment profiling offers a practical audit tool for aligning website copy with stakeholder expectations and governance communication, supporting benchmarking, and future tests linking narrative tone to investor behavior and firm performance. Full article

In our longitudinal study, we examined self-reported or explicit basic emotions, i.e., happiness, sadness, anxiety, and anger, in depressed patients during acute illness and inpatient treatment. For exploratory purposes, we also assessed implicit emotions. We analyzed how changes in emotional clarity relate to changes in emotions and depressive symptoms. A sample of depressed inpatients (n = 52) was examined at admission and on average after seven weeks of multimodal psychiatric treatment. A healthy control group (n = 52) was tested at the same time interval. Basic emotions were measured via the Differential Emotions Scale and a discrete-emotions variant of the Implicit Positive and Negative Affect Test. Emotional clarity was measured with the WEFG scales. Patients reported lower explicit happiness and heightened explicit sadness, anxiety, and anger compared to healthy controls, regardless of time of measurement. Across groups and time points, implicit happiness was greater than implicit sadness, anxiety, and anger, with no group differences. Patients’ emotional clarity improved and correlated with improvements in depressive symptoms, explicit happiness, sadness, and implicit anger. In summary, depressed patients experience heightened anxiety and anger, suggesting broader alterations of negative emotions beyond sadness. Increased emotional clarity during treatment was found to be correlated with changes in explicit and implicit affectivity. Full article

(1) Background: Virtual environments (VEs) significantly influence human emotions through various elements such as lighting, color, and terrain. While the effects of lighting and color on emotions within VEs have been extensively studied, the impact of the terrain remains underexplored. This paper addresses this gap by investigating the correlation between terrain characteristics in VEs and users’ emotional states. (2) Methods: We conducted a user study in which participants were exposed to various 3D terrains and used the Self-Assessment Manikin (SAM) to rate their emotional responses (valence, arousal, and dominance). Building on these insights, we propose MoodScape, an automated framework for emotion-informed terrain generation that significantly reduces the need for extensive expertise and manual effort. In the current implementation, continuous SAM valence–arousal targets are discretised into four quadrant-based affect/terrain classes, and this discrete class label conditions DH-CVAE-GAN terrain synthesis. MoodScape designs a generative adversarial network (GAN) architecture called DH-CVAE-GAN, which integrates a dual-head conditional variational autoencoder as the generator alongside a discriminator network to ensure effective and realistic terrain generation. The DH-CVAE-GAN is trained on a satellite-derived digital elevation model (DEM) dataset, which helps the generated terrains reflect realistic geographic patterns. (3) Results: Quantitative and qualitative evaluations on our study sample suggest that MoodScape can generate terrains whose perceived affective tone is broadly consistent with the specified affect-class inputs, indicating potential applications in gaming and exploratory therapeutic Virtual Reality, while formal clinical efficacy remains in future work. Full article

Facial features hold information about a person’s emotions, motor function, or genetic defects. Since most current mobile devices are capable of real-time face detection using cameras and depth sensors, real-time facial analysis can be utilized in several mobile use cases. Understanding the real-time emotion recognition capabilities of device sensors and frameworks is vital for developing new, valid applications. Therefore, we evaluated on-device emotion recognition using Apple’s ARKit on an iPhone 14 Pro. A native app elicited 36 blend shape-specific movements and 7 discrete emotions from $N=31$ healthy adults. Per frame, standardized ARKit blend shapes were classified using a prototype-based cosine similarity metric; performance was summarized as accuracy and area under the receiver operating characteristic curves. Cosine similarity achieved an overall accuracy of $68.3\%$, exceeding the mean of three human raters ($58.9\%$; $+9.4$ percentage points, ≈16% relative). Per-emotion accuracy was highest for joy, fear, sadness, and surprise, and competitive for anger, disgust, and contempt. AUCs were ≥0.84 for all classes. The method runs in real time on-device using only vector operations, preserving privacy and minimizing compute. These results indicate that a simple, interpretable cosine-similarity classifier over ARKit blend shapes delivers human-comparable, real-time facial emotion recognition on commodity hardware, supporting privacy-preserving mobile applications. Full article

This research addresses the challenge of inferring complex psychological states, including stress, fatigue, anxiety, cognitive load, and boredom, from facial expressions. We propose an interpretable, literature-informed emotion-weighting methodology that transforms the eight-emotion probability outputs of facial emotion recognition models into continuous estimates of these five psychological states using weights derived from the Valence–Arousal framework, providing a principled bridge between discrete emotion predictions and higher-level affective constructs. The proposed formulation is evaluated across six representative deep learning architectures—a baseline CNN (ResNet-50), a modern CNN (ConvNeXt), a hybrid attention-based model (DDAMFN), and three Transformer-based models (ViT, BEiT, and Swin). Our results demonstrate that strong performance on discrete FER tasks does not directly translate to consistent behavior in complex state inference; instead, architectures capable of preserving subtle and distributed affective cues yield more stable and interpretable state estimates, with DDAMFN and Vision Transformer models exhibiting the most consistent performance across the evaluated psychological states. These findings highlight the central role of the proposed emotion-weighting formulation and the importance of architecture selection beyond categorical accuracy in complex affective state analysis. Full article

How different negative emotions influence cognitive processes in general, and arithmetic in particular, remains poorly understood, and even less is known about how these effects change with aging in adulthood. The present study investigated whether disgust, fear, and sadness exert distinct effects on strategy selection and execution in arithmetic, and whether these effects vary across the adult lifespan. Young and older participants were asked to choose between two strategies (Experiment 1) and to execute instructed strategies (Experiment 2) to estimate the products of two-digit multiplication problems. Interestingly, how fear, disgust, and sadness influence strategy selection and strategy execution differed in young and older adults. Discrete negative emotions differentially influenced strategic aspects of arithmetic performance in young adults, whereas none modulated strategy selection or execution in older adults. These findings have important implications for furthering our understanding of emotion–cognition interactions as well as age-related changes in these interactions. Full article

Background/Objectives: Complex emotions in daily life often arise as mixtures of basic emotions, but most emotion-recognition systems still target a small set of discrete states and rely on contact-based sensing. This study aimed (1) to examine whether four compound emotions—Positive Surprise, Negative Surprise, Positive Sadness, and Negative Sadness—defined by valence direction within basic emotion categories can be differentiated using heart rate variability (HRV), and (2) to evaluate the feasibility of a camera-based contactless system (Deep Health Vision System, DHVS) by comparing it with a reference chest-strap device (Polar H10). Methods: Ten healthy adults viewed video clips designed to induce the four complex emotions. HRV was recorded simultaneously using Polar H10 and a webcam-based rPPG implementation of DHVS. Two-minute baseline and during-stimulus segments were extracted, and change rates of standard HRV indices were computed. After each stimulus, participants reported Valence, Arousal, Dominance, and proportional basic-emotion composition. Statistical analyses examined within-condition HRV changes, associations between HRV and self-reports, differences across emotion/valence conditions, and concordance between DHVS and Polar H10. Results: Self-reports confirmed distinct affective profiles for the four compound emotions. Positive and Negative Surprise were associated with heart rate reduction, while Positive Sadness showed reduced total power; Negative Sadness yielded heterogeneous but nonsignificant HRV changes. Specific HRV indices demonstrated condition-dependent correlations with Valence, Arousal, and Dominance. LF/HF changes were more sensitive to emotion category (Surprise vs. Sadness), whereas total power changes were more sensitive to valence (positive vs. negative). DHVS partially reproduced Polar H10 HRV patterns, with clearer concordance under positive-valence conditions. Conclusions: HRV captures distinct autonomic signatures of complex emotions defined by valence direction and shows meaningful links with subjective affective evaluations. LF/HF and total power provide complementary information on emotion category and valence-related autonomic reactivity, supporting indicator-specific modeling strategies. DHVS shows preliminary feasibility as a contactless HRV sensing platform for complex emotion recognition, warranting further validation with larger samples and more robust rPPG processing. Full article

Research in the biomedical field often faces challenges due to the scarcity and high cost of data, which significantly limit the development and application of machine learning models. This paper introduces a data-centric AI framework for EEG-based emotion recognition that emphasizes improving data quality rather than model complexity. Instead of proposing a deep architecture, we demonstrate how participant-guided noise filtering combined with systematic data augmentation can substantially enhance system performance across multiple classification settings: binary (high vs. low arousal), four-quadrant emotions, and seven discrete emotions. Using the SEED-VII dataset, we show that these strategies consistently improve accuracy and F1 scores, achieving competitive or superior performance compared to more sophisticated published models. The findings highlight a practical and reproducible pathway for advancing biomedical AI systems, showing that prioritizing data quality over architectural novelty yields robust and generalizable improvements in emotion recognition. Full article

Growing concerns about sustainability highlight the need to understand not only rational but also neurocognitive mechanisms that shape consumer decisions. This study examines how discrete emotions—such as empathy, moral satisfaction, and responsibility—interact with reflective cognitive control to influence green purchase intention, with neural loyalty functioning as a mediating mechanism. Grounded in dual-process theory, the proposed model is empirically tested through PLS-SEM using data from 276 consumers in Romania, Poland, and the Czech Republic, actively engaged with ecological products. The results demonstrate that both emotional and cognitive dimensions significantly predict purchase intention, while neural loyalty partially mediates these relationships, transforming temporary reactions into stable behavioral patterns. These findings suggest that bridging the intention–behaviour gap in sustainable consumption requires the integration of affective engagement, rational evaluation, and loyalty consolidation. The study contributes to sustainable marketing literature by positioning neurocognitive drivers as key antecedents of pro-environmental behaviour and by offering practical insights for designing interventions that effectively convert green intentions into consistent actions. All structural relationships were statistically significant (p < 0.05), confirming the robustness of the proposed model. Full article

While sadness and anger are distinct emotional states, the cognitive traits that differentiate people prone to one versus the other are not well understood. This research tested whether the cognitive signatures of state emotions extend to the trait level. Across two studies, we developed and validated a new Trait Sadness Scale (TSS) and used it to compare the cognitive responses of a sadness-prone group (high sadness, low anger) and an anger-prone group (high anger, low sadness) to ambiguous negative events. Contrary to predictions from state emotion theories, the groups did not differ in their causal attribution patterns (i.e., who they blamed). However, key cognitive differences did emerge: the sadness-prone group reported significantly greater helplessness, an effect specific to interpersonal contexts, and appraised the causes of negative events as more stable and global. These findings reveal a dissociation between state- and trait-level cognition, suggesting that emotional dispositions are differentiated not by simple patterns of blame, but by a more complex interplay of context-dependent appraisals of control and a pessimistic explanatory style. Full article

This study examined the influence of brand presence and discrete emotions on consumer acceptance and purchase intent of commercial chicken noodle soups. A total of 324 evaluations across three soup categories (chunky, low-sodium, condensed) were conducted under blind and unblinded conditions using a 42-term emotion lexicon. Brand presence did not exert moderate-to-large effects, though subtle brand-specific differences cannot be excluded. Instead, three emotions, “satisfied,” “disgusted,” and, for condensed soups, “bored,” emerged as the strongest predictors, together explaining a substantial proportion of variance in liking and purchase intent. Many other positive emotions clustered around “satisfied,” highlighting a parsimonious set of dominant drivers. Quiet positive emotions such as contentment, peacefulness, and warmth consistently aligned with both acceptance and purchase intent. These findings extend prior research by showing that consumer responses consolidate around a limited set of emotions, underscoring that evoking subtle, self-focused positive feelings may be more effective in comfort food marketing and product development than reliance on brand identity or nostalgia. Full article