A statistical feature extraction for deep speech emotion recognition in a bilingual scenario

Sekkate, Sara; Adib, Abdellah

doi:10.1007/s11042-022-14051-z

Cited by 10 publications

(4 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Certain features or activities lend themselves to particular methodologies more than others, and some methodologies may call for significant amounts of labelled data. Authors have proposed two methods: (a) based on a statistical-based parameterization framework for representing the speech through a fixed-length vector and (b) a deep learning approach that combines three convolutional neural networks architectures (1) . Their results achieved 87.08% and 83.90% for RAVDESS and EMOVO datasets respectively.…”

Section: Review Of Previous Work On Sermentioning

confidence: 99%

See 1 more Smart Citation

Feature Extraction of Assamese Speech Based One Motion Analysis

Gogoi,

Sharma,

Bordoloi

et al. 2023

IJST

View full text Add to dashboard Cite

Objectives:The present work aims to investigate the recognition of emotion from Assamese speech. Methods: This work presents a method based on the Gaussian Mixture Model (GMM) classifier and Mel-frequency cepstral coefficients (MFCC) as feature extraction technique for emotion recognition from Assamese speeches. Findings: We have conducted experiments considering different emotions: Angry, Happy, Neutral and Sad. The speech emotion recognition system database is the emotional speech samples collected manually from 20 speakers and some standard samples available on the internet. The speakers are from different districts of Assam and use different dialects of the Assamese language, such as Western (Kamrupi), Central, and Eastern. They fall under the age group of 18-26 years. The field survey consists of recordings done at Dibrugarh University and outside the campus. After the GMM training and testing process, the accuracy we obtained is 51.25%. The experiments confirmed that angry and happy emotions have high energy in the higher frequency region. In contrast, neutral and sad emotions have low energy in the higher frequency region. Novelty: This work will help predict the attitudes and actions of different speakers based on their manner of speaking. In addition, the present work will also help in other aspects of human-machine interaction in our daily life. The Assamese emotional speech database used in the work is self-collected from different dialect groups to understand the variability of emotions in dialectal perspective.

show abstract

Section: Review Of Previous Work On Sermentioning

confidence: 99%

“…Speech emotion can be considered as a similar type of stress on every sound event during the speech. An emotive speech describes specific prosody in speech (1) . A language's prosodic rules change throughout time as a community's culture does.…”

Section: Introductionmentioning

confidence: 99%

Feature Extraction of Assamese Speech Based One Motion Analysis

Gogoi,

Sharma,

Bordoloi

et al. 2023

IJST

View full text Add to dashboard Cite

show abstract

“…Sekkate et al (2023) [14] introduced a statistical-based technique for speech emotion recognition. The researchers first converted the speech signals into MFCC where the mean values of MFCC were calculated as features.…”

mentioning

confidence: 99%

MaxMViT-MLP: Multiaxis and Multiscale Vision Transformers Fusion Network for Speech Emotion Recognition

Ong,

Lee,

Lim

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Vision Transformers, known for their innovative architectural design and modeling capabilities, have gained significant attention in computer vision. This paper presents a dual-path approach that leverages the strengths of the Multi-Axis Vision Transformer (MaxViT) and the Improved Multiscale Vision Transformer (MViTv2). It starts by encoding speech signals into Constant-Q Transform (CQT) spectrograms and Mel Spectrograms with Short-Time Fourier Transform (Mel-STFT). The CQT spectrogram is then fed into the MaxViT model, while the Mel-STFT is input to the MViTv2 model to extract informative features from the spectrograms. These features are integrated and passed into a Multilayer Perceptron (MLP) model for final classification. This hybrid model is named the "MaxViT and MViTv2 Fusion Network with Multilayer Perceptron (MaxMViT-MLP)." The MaxMViT-MLP model achieves remarkable results with an accuracy of 95.28% on the Emo-DB, 89.12% on the RAVDESS dataset, and 68.39% on the IEMOCAP dataset, substantiating the advantages of integrating multiple audio feature representations and Vision Transformers in speech emotion recognition.

show abstract

“…These algorithms utilize signal processing techniques, feature extraction, and machine learning to analyze acoustic properties and patterns related to emotional states. Acoustic features are extracted to capture the valence and arousal dimensions, including voice quality features [89][90][91][92]. Such AI-based visual and audio algorithms are integrated in Furhat robot, allowing it to display emotions and expressions on its face by animations that correspond with the emotional content and tone of the speech being delivered.…”

mentioning

confidence: 99%

Exploring the Potential of Social Robots for Speech and Language Therapy: A Review and Analysis of Interactive Scenarios

et al. 2023

View full text Add to dashboard Cite

The use of innovative technology in the field of Speech and Language Therapy (SLT) has gained significant attention nowadays. Despite being a promising research area, Socially Assistive Robots (SARs) have not been thoroughly studied and used in SLT. This paper makes two main contributions: firstly, providing a comprehensive review of existing research on the use of SARs to enhance communication skills in children and adolescents. Secondly, organizing the information into tables that categorize the interactive play scenarios described in the surveyed papers. The inclusion criteria for play scenarios in the tables are based only on their effectiveness for SLT proven by experimental findings. The data, systematically presented in a table format, allow readers to easily find relevant information based on various factors, such as disorder type, age, treatment technique, robot type, etc. The study concludes that the despite limited research on the use of social robots for children and adolescents with communication disorders (CD), promising outcomes have been reported. The authors discuss the methodological, technical, and ethical limitations related to the use of SARs for SLT in clinical or home environments, as well as the huge potential of conversational Artificial Intelligence (AI) as a secondary assistive technology to facilitate speech and language interventions.

show abstract

A statistical feature extraction for deep speech emotion recognition in a bilingual scenario

Cited by 10 publications

References 62 publications

Feature Extraction of Assamese Speech Based One Motion Analysis

Feature Extraction of Assamese Speech Based One Motion Analysis

MaxMViT-MLP: Multiaxis and Multiscale Vision Transformers Fusion Network for Speech Emotion Recognition

Exploring the Potential of Social Robots for Speech and Language Therapy: A Review and Analysis of Interactive Scenarios

Contact Info

Product

Resources

About