Emotional Speech Research Articles

Enhanced Speech Emotion Recognition Using Conditional-DCGAN-Based Data Augmentation

With the advancement of Artificial Intelligence (AI) and the Internet of Things (IoT), research in the field of emotion detection and recognition has been actively conducted worldwide in modern society. Among this research, speech emotion recognition has gained increasing importance in various areas of application such as personalized services, enhanced security, and the medical field. However, subjective emotional expressions in voice data can be perceived differently by individuals, and issues such as data imbalance and limited datasets fail to provide the diverse situations necessary for model training, thus limiting performance. To overcome these challenges, this paper proposes a novel data augmentation technique using Conditional-DCGAN, which combines CGAN and DCGAN. This study analyzes the temporal signal changes using Mel-spectrograms extracted from the Emo-DB dataset and applies a loss function calculation method borrowed from reinforcement learning to generate data that accurately reflects emotional characteristics. To validate the proposed method, experiments were conducted using a model combining CNN and Bi-LSTM. The results, including augmented data, achieved significant performance improvements, reaching WA 91.46% and UAR 91.61%, compared to using only the original data (WA 79.31%, UAR 78.16%). These results outperform similar previous studies, such as those reporting WA 84.49% and UAR 83.33%, demonstrating the positive effects of the proposed data augmentation technique. This study presents a new data augmentation method that enables effective learning even in situations with limited data, offering a progressive direction for research in speech emotion recognition.

Graph Neural Network-Based Speech Emotion Recognition: A Fusion of Skip Graph Convolutional Networks and Graph Attention Networks

In speech emotion recognition (SER), our research addresses the critical challenges of capturing and evaluating node information and their complex interrelationships within speech data. We introduce Skip Graph Convolutional and Graph Attention Network (SkipGCNGAT), an innovative model that combines the strengths of skip graph convolutional networks (SkipGCNs) and graph attention networks (GATs) to address these challenges. SkipGCN incorporates skip connections, enhancing the flow of information across the network and mitigating issues such as vanishing gradients, while also facilitating deeper representation learning. Meanwhile, the GAT in the model assigns dynamic attention weights to neighboring nodes, allowing SkipGCNGAT to focus on both the most relevant local and global interactions within the speech data. This enables the model to capture subtle and complex dependencies between speech segments, thus facilitating a more accurate interpretation of emotional content. It overcomes the limitations of previous single-layer graph models, which were unable to effectively represent these intricate relationships across time and in different speech contexts. Additionally, by introducing a pre-pooling SkipGCN combination technique, we further enhance the ability of the model to integrate multi-layer information before pooling, improving its capacity to capture both spatial and temporal features in speech. Furthermore, we rigorously evaluated SkipGCNGAT on the IEMOCAP and MSP-IMPROV datasets, two benchmark datasets in SER. The results demonstrated that SkipGCNGAT consistently achieved state-of-the-art performance. These findings highlight the effectiveness of the proposed model in accurately recognizing emotions in speech, offering valuable insights and a solid foundation for future research on capturing complex relationships within speech signals for emotion recognition.

CNN-Based Models for Emotion and Sentiment Analysis Using Speech Data

The study aims to present an in-depth Sentiment Analysis (SA) grounded by the presence of emotions in the speech signals. Nowadays, all kinds of web-based applications ranging from social media platforms and video-sharing sites to e-commerce applications provide support for Human–Computer Interfaces (HCIs). These media applications allow users to share their experiences in all forms such as text, audio, video, GIF, and so on. The most natural and fundamental form of expressing oneself is through speech. Speech-Based Sentiment Analysis (SBSA) is the task of gaining insights into speech signals. It aims to classify the statement as neutral, negative, or positive. On the other hand, Speech Emotion Recognition (SER) categorizes speech signals into the following emotions: disgust, fear, sadness, anger, happiness, and neutral. It is necessary to recognize the sentiments along with the profoundness of the emotions in the speech signals. To cater to the above idea, a methodology is proposed defining a text-oriented SA model using the combination of CNN and Bi-LSTM techniques along with an embedding layer, applied to the text obtained from speech signals; achieving an accuracy of 84.49%. Also, the proposed methodology suggests an Emotion Analysis (EA) model based on the CNN technique highlighting the type of emotion present in the speech signal with an accuracy measure of 95.12%. The presented architecture can also be applied to different other domains like product review systems, video recommendation systems, education, health, security, and so on.

Research on Speech Emotion Recognition Method Based on ResSE_CNN1D

This article examines a speech emotion recognition (SER) technique based on the enhanced one-dimensional convolution neural network ResSE_CNN1D. With the artificial intelligence developing rapidly, SER has a profound impact in many areas. The model of this article is used to extract the characteristics of the input data through opensmile, and is sent into the ResSE_CNN1D model, which is eventually classified by the softmax activation function and obtains the final results. The key to this model is efficient learning of decimal sets and the rapid deployment in resource-constrained environments. The ResSE_CNN1D model improves the performance of the model by adding the residual connection and the SE module on the basis of cnn1d. This increasing the accuracy of the recognition and preventing the fitting problem. After the model was created, the study adopted the audio sampling and training of the casia data concentration. The final accuracy was 0.900, which increased the accuracy of 2.9 percent compared to the cnn1d method. And by the analysis of the relationship diagram of the confusion matrix and the accuracy and loss rate relative to the number of training, the model has a high robustness and effectively prevents the appearance of the fitting problem. And it also can achieve high precision and achieve a lightweight goal relative to less training.

Emotion Classification from Speech Waveform Using Machine Learning and Deep Learning Techniques

Emotions play a key role in determining the human mental state and indirectly express an individual’s well- being. A speech emotion recognition system can extract a person’s emotions from his/her speech inputs. There are some universal emotions such as anger, disgust, fear, happiness, pleasantness, sadness and neutral. These emotions are of significance especially in a situation like the Covid pandemic, when the aged or sick are vulnerable to depression. In the current paper, we examined various classification models with finite computational strength and resources in order to determine the emotion of a person from his/her speech. Speech prosodic features like pitch, loudness, and tone of speech, and work spectral features such as Mel Frequency Capstral Coefficients (MFCCs) of the voice were used to analyze the emotions of a person. Although sequence to sequence state of the art models for speech detection that offer high levels of accuracy and precision are currently in use, the computational needs of such approaches are high and inefficient. Therefore, in this work, we emphasised analysis and comparison of different classification algorithms such as multi layer perceptron, decision tree, support vector machine, and deep neural networks such as convolutional neural network and long short term memory. Given an audio file, the emotions that were exhibited by the speaker were recognized using machine learning and deep learning techniques. A comparative study was performed to identify the most appropriate algorithms that could be used to recognize emotions. Based on the experiment results, the MLP classifier and convolutional neural network model offered better accuracy with smaller variations when compared with other models used for the study.

Classification of Infant Crying Sounds Using SE-ResNet-Transformer

Recently, emotion analysis has played an important role in the field of artificial intelligence, particularly in the study of speech emotion analysis, which can help understand one of the most direct ways of human emotional communication—speech. This study focuses on the emotion analysis of infant crying. Within cries lies a variety of information, including hunger, pain, and discomfort. This paper proposes an improved classification model using ResNet and transformer. It utilizes modified Mel-frequency cepstral coefficient Mel-frequency cepstral coefficient (MFCC) features obtained through feature engineering from infant cries and integrates SE attention mechanism modules into residual blocks to enhance the model’s ability to adjust channel weights. The proposed method achieved 93% accuracy rate in experiments, offering advantages of shorter training time and higher accuracy compared to other traditional models. It provides an efficient and stable solution for infant cry classification.

ESERNet: Learning spectrogram structure relationship for effective speech emotion recognition with swin transformer in classroom discourse analysis

Speech emotion recognition (SER) has received increased attention due to its extensive applications in many fields, especially in the analysis of teacher-student dialogue in classroom environment. It can help teachers to better learn about students’ emotions and thereby adjust teaching activities. However, SER has faced several challenges, such as the intrinsic ambiguity of emotions and the complex task of interpreting emotions from speech in noisy environments. These issues can result in reduced recognition accuracy due to a focus on less relevant or insignificant features. To address these challenges, this paper presents ESERNet, a Transformer-based model designed to effectively extract crucial clues from speech data by capturing both pivotal cues and long-range relationships in speech signal. The major contribution of our approach is a two-pathway SER framework. By leveraging the Transformer architecture, ESERNet captures long-range dependencies within speech mel-spectrograms, enabling a refined understanding of the emotional cues embedded in speech signals. Extensive experiments were conducted on the IEMOCAP and EmoDB datasets, the results show that ESERNet achieves state-of-the-art performance in SER and outperforms existing methods by effectively leveraging critical clues and capturing long-range dependencies in speech data. These results highlight the effectiveness of the model in addressing the complex challenges associated with SER tasks.

DSTM: A transformer-based model with dynamic-static feature fusion in speech emotion recognition

With the support of multi-head attention, the Transformer shows remarkable results in speech emotion recognition. However, existing models still suffer from the inability to accurately locate important regions in semantic information at different time scales. To address this problem, we propose a Transformer-based network model for dynamic-static feature fusion, composed of a locally adaptive multi-head attention module and a global static attention module. The locally dynamic multi-head attention module adapts the attention window sizes and window centers of the different regions through speech samples and learnable parameters, enabling the model to adaptively discover and pay attention to valuable information embedded in speech. The global static attention module enables the model to use each element in the sequence fully and learn critical global feature information by establishing connections over the entire input sequence. We also use the data mixture training method to train our model and introduce the CENTER LOSS function to supervise the training of the model, which can better speed up the fitting speed of the model and alleviate the sample imbalance problem to a certain extent. This method achieved good performance on the IEMOCAP and MELD datasets, proving that our proposed model structure and method have better accuracy and robustness.

Graph-based multi-Feature fusion method for speech emotion recognition

Graph-based multi-Feature fusion method for speech emotion recognition

Data augmentation using a 1D-CNN model with MFCC/MFMC features for speech emotion recognition

Speech emotion recognition (SER) is attractive in several domains, such as automated translation, call centres, intelligent healthcare, and human–computer interaction. Deep learning models for emotion identification need considerable labelled data, which is only sometimes available in the SER industry. A database needs enough speech samples, good features, and a better classifier to identify emotions efficiently. This study uses data augmentation to enhance the amount of input voice samples and address the data shortage issue. The database capacity increases by adding white noise to the speech signals by data augmentation. In this work, the Mel-frequency Cepstral Coefficient (MFCC) and Mel-frequency Magnitude Coefficient (MFMC) features, along with a one-dimensional convolutional neural network (1D-CNN), are used to classify speech emotions. The datasets utilized to estimate the model's enactment were AESDD, CAFE, EmoDB, IEMOCAP, and MESD. The data augmentation with the 1D-CNN (MFMC) model performed best, with an average accuracy of 99.2% for AESDD, 99.5% for CAFE, 97.5% for EmoDB, 92.4% for IEMOCAP and 96.9% for the MESD database. The proposed 1D-CNN (MFMC) with data augmentation outperforms the 1D-CNN (MFCC) without data augmentation in emotion recognition.

Multimodal speech emotion recognition optimization using genetic algorithm

Multimodal speech emotion recognition optimization using genetic algorithm

Cross-feature fusion speech emotion recognition based on attention mask residual network and Wav2vec 2.0

Speech Emotion Recognition (SER) has received widespread attention as a crucial way for understanding human emotional states. However, the impact of irrelevant information on speech signals and data sparsity limit the development of SER system. To address these issues, this paper proposes a framework that incorporates the Attentive Mask Residual Network (AM-ResNet) and the self-supervised learning model Wav2vec 2.0 to obtain AM-ResNet features and Wav2vec 2.0 features respectively, together with a cross-attention module to interact and fuse these two features. The AM-ResNet branch mainly consists of maximum amplitude difference detection, mask residual block, and an attention mechanism. Among them, the maximum amplitude difference detection and the mask residual block act on the pre-processing and the network, respectively, to reduce the impact of silent frames, and the attention mechanism assigns different weights to unvoiced and voiced speech to reduce redundant emotional information caused by unvoiced speech. In the Wav2vec 2.0 branch, this model is introduced as a feature extractor to obtain general speech features (Wav2vec 2.0 features) through pre-training with a large amount of unlabeled speech data, which can assist the SER task and cope with data sparsity problems. In the cross-attention module, AM-ResNet features and Wav2vec 2.0 features are interacted with and fused to obtain the cross-fused features, which are used to predict the final emotion. Furthermore, multi-label learning is also used to add ambiguous emotion utterances to deal with data limitations. Finally, experimental results illustrate the usefulness and superiority of our proposed framework over existing state-of-the-art approaches.

MFGCN: Multimodal fusion graph convolutional network for speech emotion recognition

Speech emotion recognition (SER) is challenging owing to the complexity of emotional representation. Hence, this article focuses on multimodal speech emotion recognition that analyzes the speaker’s sentiment state via audio signals and textual content. Existing multimodal approaches utilize sequential networks to capture the temporal dependency in various feature sequences, ignoring the underlying relations in acoustic and textual modalities. Moreover, current feature-level and decision-level fusion methods have unresolved limitations. Therefore, this paper develops a novel multimodal fusion graph convolutional network that comprehensively executes information interactions within and between the two modalities. Specifically, we construct the intra-modal relations to excavate exclusive intrinsic characteristics in each modality. For the inter-modal fusion, a multi-perspective fusion mechanism is devised to integrate the complementary information between the two modalities. Substantial experiments on the IEMOCAP and RAVDESS datasets and experimental results demonstrate that our approach achieves superior performance.

Improved ShuffleNet V2 network with attention for speech emotion recognition

This paper enhanced human-robot interaction (HRI) towards interacting with humans through a speech-emotion recognition algorithm based on the improved ShuffleNet V2 network as the means of communication. When recognizing speech emotion, previous studies typically analyze speech signals solely from the time or frequency domain, resulting in the loss of detailed information. Our algorithm initially converts speech data into an acoustic spectrogram to address this limitation, preserving rich original information across both time and frequency domains. This approach optimizes feature extraction by leveraging the capabilities of deep convolutional networks to capture local changes in images. Furthermore, we integrated the attention mechanism from ECA-Net into the ShuffleNet V2 model. This incorporation enhances the extraction of significant features, amplifying the expression of features strongly correlated with the emotional state conveyed in speech while suppressing irrelevant ones. In addition to these enhancements, we replace the ReLU activation function with Hardswish to deepen the neural network, thereby improving feature information without compromising model accuracy. Finally, the training stage incorporates the hot restart Cosine Annealing Learning Rate to optimize the network further. Simulation experiments are conducted on the enterface'05 dataset to evaluate the accuracy of the proposed algorithm. The experimental results show that the recognition rate of the method is higher than that of the existing algorithm in terms of the effectiveness of HRI.

Extending speech emotion recognition systems to non-prototypical emotions using mixed-emotion model

In the conventional approach to speech emotion recognition (SER), the classifier is usually trained on acted emotional speech data to predict individual basic emotions. In this work, we extend the SER systems with the realistic assumption of the coexistence of multiple basic emotions in an utterance. We utilize the MSP-Podcast database for developing the SER system, which contains spontaneous speech utterances. From the primary and secondary emotion annotations of this database, we organize six prototypical and seven non-prototypical emotions. We then propose the mixed-emotion model and express the non-prototypical emotions as a linear combination of prototypical emotions. The combination weights of the mixed-emotion model are computed using a normalized dominance scale based algorithm inspired by the integration of basic emotion theory and dimensional emotion theory in human psychology. We first train a prototypical SER model using the ECAPA-TDNN architecture. The softmax predictions from this model serve as emotion profile inputs to the mixed-emotion model, which then predicts the non-prototypical emotions. Assuming the coexistence of multiple emotions, we only apply the utterances with uniform emotion profiles to the mixed-emotion model. The developed system continues with the conventional SER model if the emotion profile tends to delta function owing to the probable occurrence of a single prototypical emotion. We develop the proposed mixed-emotion model based SER framework using MFCC and wav2vec 2.0 extracted features. Further, we show that due to human perception variations, there exist prominent annotation variations in the non-prototypical emotion ground truths. To address that, we extend the supervised evaluation protocols in four different formulations that capture the subjective variability at different levels. The proposed system shows a best-case performance improvement of 7.10% and 8.39% over the conventional prototypical SER model for the MFCC and wav2vec 2.0 features, respectively.

Multimodal fusion: A study on speech-text emotion recognition with the integration of deep learning

Recognition of various human emotions holds significant value in numerous real-world scenarios. This paper focuses on the multimodal fusion of speech and text for emotion recognition. A 39-dimensional Mel-frequency cepstral coefficient (MFCC) was used as a feature for speech emotion. A 300-dimensional word vector obtained through the Glove algorithm was used as the feature for text emotion. The bidirectional gate recurrent unit (BiGRU) method in deep learning was added for extracting deep features. Subsequently, it was combined with the multi-head self-attention (MHA) mechanism and the improved sparrow search algorithm (ISSA) to obtain the ISSA-BiGRU-MHA method for emotion recognition. It was validated on the IEMOCAP and MELD datasets. It was found that MFCC and Glove word vectors exhibited superior recognition effects as features. Comparisons with the support vector machine and convolutional neural network methods revealed that the ISSA-BiGRU-MHA method demonstrated the highest weighted accuracy and unweighted accuracy. Multimodal fusion achieved weighted accuracies of 76.52 %, 71.84 %, 66.72 %, and 62.12 % on the IEMOCAP, MELD, MOSI, and MOSEI datasets, suggesting better performance than unimodal fusion. These results affirm the reliability of the multimodal fusion recognition method, showing its practical applicability.

CENN: Capsule-enhanced neural network with innovative metrics for robust speech emotion recognition

Speech emotion recognition (SER) plays a pivotal role in enhancing Human-computer interaction (HCI) systems. This paper introduces a groundbreaking Capsule-enhanced neural network (CENN) that significantly advances the state of SER through a robust and reproducible deep learning framework. The CENN architecture seamlessly integrates advanced components, including Multi-head attention (MHA), residual module, and capsule module, which collectively enhance the model's capacity to capture both global and local features essential for precise emotion classification. A key contribution of this work is the development of a comprehensive reproducibility framework, featuring novel metrics: General learning reproducibility (GLR) and Correct learning reproducibility (CLR). These metrics, alongside their fractional and perfect variants, offer a multi-dimensional evaluation of the model's consistency and correctness across multiple executions, thereby ensuring the reliability and credibility of the results. To tackle the persistent challenge of overfitting in deep learning models, we propose an innovative overfitting metric that considers the intricate relationship between training and testing errors, model complexity, and data complexity. This metric, in conjunction with the newly introduced generalization and robustness metrics, provides a holistic assessment of the model's performance, guiding the application of regularization techniques to maintain generalizability and resilience. Extensive experiments conducted on benchmark SER datasets demonstrate that the CENN model not only surpasses existing approaches in terms of accuracy but also sets a new benchmark in reproducibility. This work establishes a new paradigm for deep learning model development in SER, underscoring the vital importance of reproducibility and offering a rigorous framework for future research.

A Combined CNN Architecture for Speech Emotion Recognition.

Emotion recognition through speech is a technique employed in various scenarios of Human-Computer Interaction (HCI). Existing approaches have achieved significant results; however, limitations persist, with the quantity and diversity of data being more notable when deep learning techniques are used. The lack of a standard in feature selection leads to continuous development and experimentation. Choosing and designing the appropriate network architecture constitutes another challenge. This study addresses the challenge of recognizing emotions in the human voice using deep learning techniques, proposing a comprehensive approach, and developing preprocessing and feature selection stages while constructing a dataset called EmoDSc as a result of combining several available databases. The synergy between spectral features and spectrogram images is investigated. Independently, the weighted accuracy obtained using only spectral features was 89%, while using only spectrogram images, the weighted accuracy reached 90%. These results, although surpassing previous research, highlight the strengths and limitations when operating in isolation. Based on this exploration, a neural network architecture composed of a CNN1D, a CNN2D, and an MLP that fuses spectral features and spectogram images is proposed. The model, supported by the unified dataset EmoDSc, demonstrates a remarkable accuracy of 96%.

Emotion Recognition Through Analysis of Speech – A Review

The feature extraction is very important for emotion recognition through speech. There are several approaches when dealing with emotion recognition. In this paper, we present different feature extraction approaches as well as different models used to differentiate between a neutral speech versus an emotional speech sample. This research is instrumental for the digitization and preservation of cultural heritage, as it allows us to capture and analyze the emotional nuances in historical audio recordings, ensuring their accurate representation for future generations. We have selected two works consisting of a total of four different methods for emotion recognition. In the first paper by Jacob (2017), we look at Decision tree and Logistic Regression. Decision tree attains an 84.45% accuracy on the test class whereas logistic regression is able to achieve an accuracy of 66.85% after stepwise regression. These methods contribute to the digital archiving of cultural heritage by providing robust tools for analyzing and preserving the emotional content of spoken artifacts. In another paper by Bhatti et all. (2004), sequential forward selection (SFS) was used to create subsets from the given features and relevance of the subsets of features. General regression neural network was used to evaluate the accuracy which was found to be 80.69%. As a complementary purpose, modular neural network was performed with an accuracy of 83.31% with the same dataset. These techniques enhance our ability to maintain the integrity and emotional depth of cultural heritage recordings in digital archives.

Fusion of PCA and ICA in Statistical Subset Analysis for Speech Emotion Recognition.

Speech emotion recognition is key to many fields, including human-computer interaction, healthcare, and intelligent assistance. While acoustic features extracted from human speech are essential for this task, not all of them contribute to emotion recognition effectively. Thus, reduced numbers of features are required within successful emotion recognition models. This work aimed to investigate whether splitting the features into two subsets based on their distribution and then applying commonly used feature reduction methods would impact accuracy. Filter reduction was employed using the Kruskal-Wallis test, followed by principal component analysis (PCA) and independent component analysis (ICA). A set of features was investigated to determine whether the indiscriminate use of parametric feature reduction techniques affects the accuracy of emotion recognition. For this investigation, data from three databases-Berlin EmoDB, SAVEE, and RAVDES-were organized into subsets according to their distribution in applying both PCA and ICA. The results showed a reduction from 6373 features to 170 for the Berlin EmoDB database with an accuracy of 84.3%; a final size of 130 features for SAVEE, with a corresponding accuracy of 75.4%; and 150 features for RAVDESS, with an accuracy of 59.9%.