Multimodal Emotion Research Articles

Multimodal Emotion Recognition Based on Facial Expressions, Speech, and Body Gestures

The research of multimodal emotion recognition based on facial expressions, speech, and body gestures is crucial for oncoming intelligent human–computer interfaces. However, it is a very difficult task and has seldom been researched in this combination in the past years. Based on the GEMEP and Polish databases, this contribution focuses on trimodal emotion recognition from facial expressions, speech, and body gestures, including feature extraction, feature fusion, and multimodal classification of the three modalities. In particular, for feature fusion, two novel algorithms including supervised least squares multiset kernel canonical correlation analysis (SLSMKCCA) and sparse supervised least squares multiset kernel canonical correlation analysis (SSLSMKCCA) are presented, respectively, to carry out efficient facial expression, speech, and body gesture feature fusion. Different from the traditional multiset kernel canonical correlation analysis (MKCCA) algorithms, our SLSKMCCA algorithm is a supervised version and is based on the least squares form. The SSLSKMCCA algorithm is implemented by the combination of SLSMKCCA and a sparse item (L1 Norm). Moreover, two effective solving algorithms for SLSMKCCA and SSLSMKCCA are presented in addition, which use the alternated least squares and augmented Lagrangian multiplier methods, respectively. The extensive experimental results on the popular public GEMEP and Polish databases show that the recognition rate of multimodal emotion recognition is superior to bimodal and monomodal emotion recognition on average, and our presented SLSMKCCA and SSLSMKCCA fusion methods both obtain very high recognition rates, especially for the SSLSMKCCA fusion method.

Multiplex graph aggregation and feature refinement for unsupervised incomplete multimodal emotion recognition

Multimodal Emotion Recognition (MER) involves integrating information of various modalities, including audio, visual, text and physiological signals, to comprehensively grasp human sentiments, which has emerged as a vibrant area within human–computer interaction. Researchers have developed many methods for this task, but many of these methods rely on labeled supervised learning and struggle to address the issue of missing some modalities of data. To address these issues, we propose a Multiplex Graph Aggregation and Feature Refinement framework for unsupervised incomplete MER, comprising four modules: Completion, Aggregation, Refinement, and Embedding. Specifically, we first capture the correlation information between samples using the graph structures, which aids in the completion of missing data and the multiplex aggregation of multimodal data. Then, we perform refinement operations on the aggregated features as well as alignment and enhancement operations on the embedding features to obtain the fused feature representations, which are consistent, highly separable and conducive to emotion recognition. Experimental results on multimodal emotion recognition datasets demonstrate that our method achieves state-of-the-art performance among unsupervised methods, validating its effectiveness.

EAV: EEG-Audio-Video Dataset for Emotion Recognition in Conversational Contexts

Understanding emotional states is pivotal for the development of next-generation human-machine interfaces. Human behaviors in social interactions have resulted in psycho-physiological processes influenced by perceptual inputs. Therefore, efforts to comprehend brain functions and human behavior could potentially catalyze the development of AI models with human-like attributes. In this study, we introduce a multimodal emotion dataset comprising data from 30-channel electroencephalography (EEG), audio, and video recordings from 42 participants. Each participant engaged in a cue-based conversation scenario, eliciting five distinct emotions: neutral, anger, happiness, sadness, and calmness. Throughout the experiment, each participant contributed 200 interactions, which encompassed both listening and speaking. This resulted in a cumulative total of 8,400 interactions across all participants. We evaluated the baseline performance of emotion recognition for each modality using established deep neural network (DNN) methods. The Emotion in EEG-Audio-Visual (EAV) dataset represents the first public dataset to incorporate three primary modalities for emotion recognition within a conversational context. We anticipate that this dataset will make significant contributions to the modeling of the human emotional process, encompassing both fundamental neuroscience and machine learning viewpoints.

Correlation mining of multimodal features based on higher-order partial least squares for emotion recognition in conversations

In fields requiring an understanding of emotions, such as digital human interaction and public opinion analysis, achieving a dependable and interpretable model for mining correlations among multimodal features remains a primary objective. However, current deep learning methods often lack transparency and suffer from low interpretability. To address these challenges, we propose a novel Correlation Mining method based on Higher-Order Partial Least Squares (HOPLS) for multimodal Emotion Recognition in conversations (CMHER) in this paper. CMHER innovatively combines HOPLS with transformers and Gated Recurrent Units (GRUs) to compute correlation matrices within unimodal data streams and between cross-modal sources. HOPLS projects source data into a latent space to predict target data via correlation matrix computations, eliminating the need for Graphical Processing Unit (GPU) acceleration and making it suitable for experimental and edge systems. The integration of HOPLS with deep neural networks involves preprocessing multimodal features into suitable dimensions and latent representations, followed by HOPLS computing correlation matrices for cross-modal latent vectors and final labels through optimal joint subspace approximation, which aims at the improvements of both interpretability and reliability. Additionally, a generalization error fitting module further refines the predicted correlation matrices to improve predictive capability and overall model performance. Experiments on two public datasets validate the superiority of our proposed method.

Multimodal emotion recognition based on a fusion of audiovisual information with temporal dynamics

AbstractIn the Human-Machine Interactions (HMI) landscape, understanding user emotions is pivotal for elevating user experiences. This paper explores Facial Expression Recognition (FER) within HMI, employing a distinctive multimodal approach that integrates visual and auditory information. Recognizing the dynamic nature of HMI, where situations evolve, this study emphasizes continuous emotion analysis. This work assesses various fusion strategies that involve the addition to the main network of different architectures, such as autoencoders (AE) or an Embracement module, to combine the information of multiple biometric cues. In addition to the multimodal approach, this paper introduces a new architecture that prioritizes temporal dynamics by incorporating Long Short-Term Memory (LSTM) networks. The final proposal, which integrates different multimodal approaches with the temporal focus capabilities of the LSTM architecture, was tested across three public datasets: RAVDESS, SAVEE, and CREMA-D. It showcased state-of-the-art accuracy of 88.11%, 86.75%, and 80.27%, respectively, and outperformed other existing approaches.

A Reinforcement Learning-Based Approach for Promoting Mental Health Using Multimodal Emotion Recognition

This research aims to enhance mental well-being by addressing symptoms of anxiety and depression through a personalized, culturally specific multimodal emotion prediction system. It employs an emotionally aware Reinforcement Learning (RL) agent to suggest tailored Cognitive Behavioral Therapy (CBT) activities. The study focuses on developing precise, individualized emotion prediction models using facial expressions, vocal tones, and text, and integrates these models with the RL agent for emotionally aware CBT recommendations. The mHealth approach combines deep learning models with RL, achieving accuracies of 72% for facial expressions, 73% for vocal tones, and 86% for text, all fine-tuned for the Sri Lankan context. Validation through real-world use and user feedback consistently demonstrated that each model exceeds 70% accuracy, fulfilling the objective of precise emotion prediction. A weighted algorithm was introduced to refine the emotion prediction experience and personalize forecasts across the three modalities to enhance mental well-being. The RL-enabled agent suggests CBT activities approved by mental health professionals, tailored based on predicted emotions, and delivered through the same mHealth application. The effectiveness of these interventions was assessed using the DASS-21 questionnaire, revealing significant reductions in depression scores (from 21.08 to 13.54) and anxiety scores (from 19.85 to 10.46) in the study group compared to the control group. The study concludes that integrating multimodal emotion prediction models with RL-based CBT suggestions positively impacts mental well-being and contributes to personalized mental health interventions.

AVaTER: Fusing Audio, Visual, and Textual Modalities Using Cross-Modal Attention for Emotion Recognition.

Multimodal emotion classification (MEC) involves analyzing and identifying human emotions by integrating data from multiple sources, such as audio, video, and text. This approach leverages the complementary strengths of each modality to enhance the accuracy and robustness of emotion recognition systems. However, one significant challenge is effectively integrating these diverse data sources, each with unique characteristics and levels of noise. Additionally, the scarcity of large, annotated multimodal datasets in Bangla limits the training and evaluation of models. In this work, we unveiled a pioneering multimodal Bangla dataset, MAViT-Bangla (Multimodal Audio Video Text Bangla dataset). This dataset, comprising 1002 samples across audio, video, and text modalities, is a unique resource for emotion recognition studies in the Bangla language. It features emotional categories such as anger, fear, joy, and sadness, providing a comprehensive platform for research. Additionally, we developed a framework for audio, video and textual emotion recognition (i.e., AVaTER) that employs a cross-modal attention mechanism among unimodal features. This mechanism fosters the interaction and fusion of features from different modalities, enhancing the model's ability to capture nuanced emotional cues. The effectiveness of this approach was demonstrated by achieving an F1-score of 0.64, a significant improvement over unimodal methods.

Commonalities and variations in emotion representation across modalities and brain regions

Humans express emotions through various modalities such as facial expressions and natural language. However, the relationships between emotions expressed through different modalities and their correlations with neural activities remain uncertain. Here, we aimed to unveil some of these uncertainties by investigating the similarity of emotion representations across modalities and brain regions. First, we represented various emotion categories as multi-dimensional vectors derived from visual (face), linguistic, and visio-linguistic data, and used representational similarity analysis to compare these modalities. Second, we examined the linear transferability of emotion representation from other modalities to the visual modality. Third, we compared the representational structure derived in the first step with those from brain activities across 360 regions. Our findings revealed that emotion representations share commonalities across modalities with modality-type dependent variations, and they can be linearly mapped from other modalities to the visual modality. Additionally, emotion representations in uni-modalities showed relatively higher similarity with specific brain regions, while multi-modal emotion representation was most similar to representations across the entire brain region. These findings suggest that emotional experiences are represented differently across various brain regions with varying degrees of similarity to different modality types, and that they may be multi-modally conveyable in visual and linguistic domains.

Multimodal Emotion Recognition Using Visual, Vocal and Physiological Signals: A Review

The dynamic expressions of emotion convey both the emotional and functional states of an individual’s interactions. Recognizing the emotional states helps us understand human feelings and thoughts. Systems and frameworks designed to recognize human emotional states automatically can use various affective signals as inputs, such as visual, vocal and physiological signals. However, emotion recognition via a single modality can be affected by various sources of noise that are specific to that modality and the fact that different emotion states may be indistinguishable. This review examines the current state of multimodal emotion recognition methods that integrate visual, vocal or physiological modalities for practical emotion computing. Recent empirical evidence on deep learning methods used for fine-grained recognition is reviewed, with discussions on the robustness issues of such methods. This review elaborates on the profound learning challenges and solutions required for a high-quality emotion recognition system, emphasizing the benefits of dynamic expression analysis, which aids in detecting subtle micro-expressions, and the importance of multimodal fusion for improving emotion recognition accuracy. The literature was comprehensively searched via databases with records covering the topic of affective computing, followed by rigorous screening and selection of relevant studies. The results show that the effectiveness of current multimodal emotion recognition methods is affected by the limited availability of training data, insufficient context awareness, and challenges posed by real-world cases of noisy or missing modalities. The findings suggest that improving emotion recognition requires better representation of input data, refined feature extraction, and optimized aggregation of modalities within a multimodal framework, along with incorporating state-of-the-art methods for recognizing dynamic expressions.

Multi-modal emotion recognition using tensor decomposition fusion and self-supervised multi-tasking

Multi-modal emotion recognition using tensor decomposition fusion and self-supervised multi-tasking

Multimodal graph learning with framelet-based stochastic configuration networks for emotion recognition in conversation

The multimodal emotion recognition in conversation (ERC) task presents significant challenges due to the complexity of relationships and the difficulty in achieving semantic fusion across various modalities. Graph learning, recognized for its capability to capture intricate data relations, has been suggested as a solution for ERC. However, existing graph-based ERC models often fail to address the fundamental limitations of graph learning, such as assuming pairwise interactions and neglecting high-frequency signals in semantically-poor modalities, which leads to an over-reliance on text. While these issues might be negligible in other applications, they are crucial for the success of ERC. In this paper, we propose a novel framework for ERC, namely multimodal graph learning with framelet-based stochastic configuration networks (i.e., Frame-SCN). Specifically, framelet-based stochastic configuration networks, which employ 2D directional Haar framelets to extract both low- and high-pass components, are introduced to learn the unified semantic embeddings from multimodal data, mitigating prediction biases caused by an excessive reliance on text without introducing an unnecessarily large number of parameters. Also, we develop a modality-aware information extraction module that is able to extract both general and sensitive information in a multimodal semantic space, alleviating potential noise issues. Extensive experiment results demonstrate that our proposed Frame-SCN outperforms many state-of-the-art approaches on two widely used multimodal ERC datasets.

EmotionCast: An Emotion-Driven Intelligent Broadcasting System for Dynamic Camera Switching.

Traditional broadcasting methods often result in fatigue and decision-making errors when dealing with complex and diverse live content. Current research on intelligent broadcasting primarily relies on preset rules and model-based decisions, which have limited capabilities for understanding emotional dynamics. To address these issues, this study proposed and developed an emotion-driven intelligent broadcasting system, EmotionCast, to enhance the efficiency of camera switching during live broadcasts through decisions based on multimodal emotion recognition technology. Initially, the system employs sensing technologies to collect real-time video and audio data from multiple cameras, utilizing deep learning algorithms to analyze facial expressions and vocal tone cues for emotion detection. Subsequently, the visual, audio, and textual analyses were integrated to generate an emotional score for each camera. Finally, the score for each camera shot at the current time point was calculated by combining the current emotion score with the optimal scores from the preceding time window. This approach ensured optimal camera switching, thereby enabling swift responses to emotional changes. EmotionCast can be applied in various sensing environments such as sports events, concerts, and large-scale performances. The experimental results demonstrate that EmotionCast excels in switching accuracy, emotional resonance, and audience satisfaction, significantly enhancing emotional engagement compared to traditional broadcasting methods.

Automatic movie genre classification & emotion recognition via a BiProjection Multimodal Transformer

Analyzing, manipulating, and comprehending data from multiple sources (e.g., websites, software applications, files, or databases) and of diverse modalities (e.g., video, images, audio and text) has become increasingly important in many domains. Despite recent advances in multimodal classification (MC), there are still several challenges to be addressed, such as: the combination of modalities of very diverse nature, the optimal feature engineering for each modality, as well as the semantic alignment between text and images. Accordingly, the main motivation of our research relies in devising a neural architecture that effectively processes and combines text, image, video and audio modalities, so it can offer a noteworthy performance in different MC tasks. In this regard, the Multimodal Transformer (MulT) model is a cutting-edge approach often employed in multimodal supervised tasks, which, although effective, has the problem of having a fixed architecture that limits its performance in specific tasks as well as its contextual understanding, meaning it may struggle to capture fine-grained temporal patterns in audio or effectively model spatial relationships in images. To address these issues, our research modifies and extends the MulT model in several aspects. Firstly, we focus on leveraging the Gated Multimodal Unit (GMU) module within the architecture to efficiently and dynamically weigh modalities at the instance level and to visualize the use of modalities. Secondly, to overcome the problem of vanishing and exploding gradients we focus on strategically placing residual connections in the architecture. The proposed architecture is evaluated in two different and complex classification tasks, on the one hand, the movie genre categorization (MGC) and, on the other hand, the multimodal emotion recognition (MER). The results obtained are encouraging as they indicate that the proposed architecture is competitive against state-of-the-art (SOTA) models in MGC, outperforming them by up to 2% on the Moviescope dataset, and by 1% on the MM-IMDB datasets. Furthermore, in the MER task the unaligned version of the datasets was employed, which is considerably more difficult; we improve accuracy SOTA results by up to 1% on the IEMOCAP dataset, and attained a competitive outcome on the CMU-MOSEI11Dai et al., (2021) collection, outperforming SOTA results in several emotions.

DGFN Multimodal Emotion Analysis Model Based on Dynamic Graph Fusion Network

In recent years, integrating text and image data for sentiment analysis in social networks has become a key approach. However, techniques for capturing complex cross-modal information and effectively fusing multimodal features still have shortcomings. We design a multimodal sentiment analysis model called the Dynamic Graph-Text Fusion Network (DGFN) to address these challenges. Text features are captured by leveraging the neighborhood information aggregation properties of Graph Convolutional Networks, treating words as nodes and integrating their features through their adjacency relationships. Additionally, the multi-head attention mechanism is utilized to extract rich semantic information from different subspaces simultaneously. For image feature extraction, a convolutional attention module is employed. Subsequently, an attention-based fusion module integrates the text and image features. Experimental results on the two datasets show significant improvements in sentiment classification accuracy and F1 scores, validating the effectiveness of the proposed DGFN model.

Multi-Modal Emotion Recognition Based on Wavelet Transform and BERT-RoBERTa: An Innovative Approach Combining Enhanced BiLSTM and Focus Loss Function

Multi-Modal Emotion Recognition Based on Wavelet Transform and BERT-RoBERTa: An Innovative Approach Combining Enhanced BiLSTM and Focus Loss Function

Enhancing Multimodal Emotion Analysis through Fusion with EMT Model Based on BBFN

Enhancing Multimodal Emotion Analysis through Fusion with EMT Model Based on BBFN

Multimodal Emotion Cognition Method Based on Multi-Channel Graphic Interaction

The relationship between the emotional components associated with images and text is a crucial way of multimodal emotion analysis. However, most of the present multimodel affective cognitive models simply associate the features of images and texts without thoroughly investigating their interactions, resulting in poor recognition. Therefore, a multimodel emotion cognition method based on multi-channel graphic interaction is proposed. Text context features are extracted, scene and image information is encoded, and useful features are obtained. Based on these results, the modal alignment module be applied to obtain information about affective regions and words, and then the cross-modal gating module be applied to combine the multimodel features. In addition, we tested extensively on three open datasets, achieving an accuracy of 0.8122 for the MSA-single dataset, 0.7307 for the MSA-MULTIPLE dataset, and 0.7159 for TumEmo. The results show that this method is effective for multimodal emotion detection.

Adversarial alignment and graph fusion via information bottleneck for multimodal emotion recognition in conversations

With the rapid development of social media and human–computer interaction, multimodal emotion recognition in conversations (MERC) tasks have begun to receive widespread research attention. The MERC task is to extract and fuse complementary semantic information from different modalities to classify the speaker’s emotion. However, the existing feature fusion methods usually directly map the features of other modalities into the same feature space for information fusion, which cannot eliminate the heterogeneity between different modalities and make the subsequent emotion class boundary learning more difficult. In addition, existing graph contrastive learning methods obtain consistent feature representations by maximizing mutual information between multiple views, which may lead to overfitting of the model. To tackle the above problem, we propose a novel Adversarial Alignment and Graph Fusion via Information Bottleneck for Multimodal Emotion Recognition in Conversations (AGF-IB) method. Firstly, we input video, audio, and text features into a multi-layer perceptron (MLP) to map them into separate feature spaces. Secondly, we build a generator and a discriminator for the three modal features, respectively, through adversarial representation to achieve information interaction between modalities and eliminate the heterogeneity among modalities. Thirdly, we introduce graph contrastive representation learning to capture intra-modal and inter-modal complementary semantic information and learn intra-class and inter-class boundary information of emotion categories. Furthermore, instead of maximizing the mutual information (MI) between multiple views, we use information bottleneck theory to minimize the MI between views. Specifically, we construct a graph structure for the three modal features respectively and perform contrastive representation learning on nodes with different emotions in the same modality and nodes with the same emotion in different modalities, to improve the feature representation ability of nodes. Finally, we use MLP to complete the emotional classification of the speaker. Extensive experiments show that AGF-IB can improve emotion recognition accuracy on IEMOCAP and MELD datasets. Furthermore, since AGF-IB is a general multimodal fusion and contrastive learning method, it can be applied to other multimodal tasks in a plug-and-play manner, e.g., humor detection.

End-to-end multimodal emotion recognition based on facial expressions and remote photoplethysmography signals.

Emotion is a complex physiological phenomenon, and a single modality may be insufficient for accurately determining human emotional states. This paper proposes an end-to-end multimodal emotion recognition method based on facial expressions and non-contact physiological signals. Facial expression features and remote photoplethysmography (rPPG) signals are extracted from facial video data, and a transformer-based cross-modal attention mechanism (TCMA) is used to learn the correlation between the two modalities. The results show that the accuracy of emotion recognition can be slightly improved by combining facial expressions with accurate rPPG signals. The performance is further improved with the use of TCMA, for which the binary classification accuracy of valence and arousal is 91.11% and 90.00%, respectively. Additionally, when experiments are conducted using the whole dataset, an increased accuracy of 7.31% and 4.23% for the binary classification of valence and arousal, and an improved accuracy of 5.36% for the four classifications of valence-arousal are achieved when TCMA is used in modal fusion, compared to using only facial expression modality, which fully demonstrates the effectiveness and robustness of TCMA. This method makes it possible to realize multimodal emotion recognition of facial expressions and contactless physiological signals in reality.

What multimodal components, tools, dataset and focus of emotion are used in the current research of multimodal emotion: a systematic literature review

What multimodal components, tools, dataset and focus of emotion are used in the current research of multimodal emotion: a systematic literature review