University Sétif 1 FERHAT ABBAS Faculty of Sciences
Détail de l'auteur
Auteur madjda Daikha |
Documents disponibles écrits par cet auteur
Ajouter le résultat dans votre panier Affiner la recherche
Transformer-based Deep Learning Techniques for Speech Emotion Recognition and Sentiment Analysis / madjda Daikha
Titre : Transformer-based Deep Learning Techniques for Speech Emotion Recognition and Sentiment Analysis Type de document : texte imprimé Auteurs : madjda Daikha, Auteur ; ilhem benchikh, Auteur ; Muussaoui,Abdelouhab, Directeur de thèse Année de publication : 2022 Importance : 1 vol (64 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Deep learning
Speech emotion recognition and sentiment analytics
TransformerIndex. décimale : 004 Informatique Résumé :
In this thesis, we want to propose a new classification model based on deep learning
that will be able to classify and identify different speech emotion recognition and
sentiment analytics with high accuracy that outperforms the state-of-the-art approaches.
Automatic speech emotion recognition is a very necessary activity for effective humancomputer
interaction. Emotional awareness by machines can also be used to provide
tools to humans to make them more effective.
We propose here a new Transformer-based deep learning architecture. Our approach
is tested and verified using the public dataset CREMA-D which contains over 7442
different people audio speech. A comparative study was made between different machine
and deep learning approaches and the results showed that the deep approach with
Transformer gave the best test accuracy where we obtained an accuracy up to 99.19%.Côte titre : MAI/0591 En ligne : https://drive.google.com/file/d/1PLkGumgGZz4dsf3BJ8w7uw6Q0D1qt57G/view?usp=share [...] Format de la ressource électronique : Transformer-based Deep Learning Techniques for Speech Emotion Recognition and Sentiment Analysis [texte imprimé] / madjda Daikha, Auteur ; ilhem benchikh, Auteur ; Muussaoui,Abdelouhab, Directeur de thèse . - 2022 . - 1 vol (64 f .) ; 29cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Deep learning
Speech emotion recognition and sentiment analytics
TransformerIndex. décimale : 004 Informatique Résumé :
In this thesis, we want to propose a new classification model based on deep learning
that will be able to classify and identify different speech emotion recognition and
sentiment analytics with high accuracy that outperforms the state-of-the-art approaches.
Automatic speech emotion recognition is a very necessary activity for effective humancomputer
interaction. Emotional awareness by machines can also be used to provide
tools to humans to make them more effective.
We propose here a new Transformer-based deep learning architecture. Our approach
is tested and verified using the public dataset CREMA-D which contains over 7442
different people audio speech. A comparative study was made between different machine
and deep learning approaches and the results showed that the deep approach with
Transformer gave the best test accuracy where we obtained an accuracy up to 99.19%.Côte titre : MAI/0591 En ligne : https://drive.google.com/file/d/1PLkGumgGZz4dsf3BJ8w7uw6Q0D1qt57G/view?usp=share [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0591 MAI/0591 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible