University Sétif 1 FERHAT ABBAS Faculty of Sciences
Détail de l'auteur
Auteur Abdelouaheb Moussaoui |
Documents disponibles écrits par cet auteur
Ajouter le résultat dans votre panier Affiner la recherche
Classification of Remote Sensing Images using Transfer Learning and CNN Model with Attention / Yousra Halfaya
Titre : Classification of Remote Sensing Images using Transfer Learning and CNN Model with Attention Type de document : texte imprimé Auteurs : Yousra Halfaya, Auteur ; Wafa Lamamra, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse Année de publication : 2022 Importance : 1 vol (98 f .) Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Remote sensing
Scene classificationIndex. décimale : 004 Informatique Résumé :
Classification of high-resolution remote sensing image is one of the advanced technologies
in our century. However, scientist’s efforts were made to progress various
forms of scene classification methods. In this thesis we will focus to identify Aerial
image scene classification and fire forest image classification based on Deep Learning
approaches. In addition, we will offer at the beginning a comprehensive introduction
about of the recent state-of-the-art models. Then we will be proposing two datasets.
Firstly, a wide-scale dataset, called “UCMerced” dataset which is available in the public
for remote sensing image scene classification. This dataset includes 2,100 images
representing 21 scene classes, with each class contains 100 images. Secondly we created
a single integrated forest fire dataset containing 2955 images, with 1695 forest fire
images and 1260 non-fire image . Based on this dataset, we propose four models based
on Transfert Learning models, a ConvNet-based model, a ConvNet-based model with
attention mechanism, and Transformers based model.
Finally, The performance of the considered models is evaluated by computing the
common performance measures.The result of the experimental evaluation show that
our proposed models improve scene classification on both datasets.Côte titre : MAI/0586 En ligne : https://drive.google.com/file/d/1CqlypwIur_Ypctq4H6IJob-N1Sq1ll_y/view?usp=share [...] Format de la ressource électronique : Classification of Remote Sensing Images using Transfer Learning and CNN Model with Attention [texte imprimé] / Yousra Halfaya, Auteur ; Wafa Lamamra, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse . - 2022 . - 1 vol (98 f .).
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Remote sensing
Scene classificationIndex. décimale : 004 Informatique Résumé :
Classification of high-resolution remote sensing image is one of the advanced technologies
in our century. However, scientist’s efforts were made to progress various
forms of scene classification methods. In this thesis we will focus to identify Aerial
image scene classification and fire forest image classification based on Deep Learning
approaches. In addition, we will offer at the beginning a comprehensive introduction
about of the recent state-of-the-art models. Then we will be proposing two datasets.
Firstly, a wide-scale dataset, called “UCMerced” dataset which is available in the public
for remote sensing image scene classification. This dataset includes 2,100 images
representing 21 scene classes, with each class contains 100 images. Secondly we created
a single integrated forest fire dataset containing 2955 images, with 1695 forest fire
images and 1260 non-fire image . Based on this dataset, we propose four models based
on Transfert Learning models, a ConvNet-based model, a ConvNet-based model with
attention mechanism, and Transformers based model.
Finally, The performance of the considered models is evaluated by computing the
common performance measures.The result of the experimental evaluation show that
our proposed models improve scene classification on both datasets.Côte titre : MAI/0586 En ligne : https://drive.google.com/file/d/1CqlypwIur_Ypctq4H6IJob-N1Sq1ll_y/view?usp=share [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0586 MAI/0586 Mémoire Bibliothéque des sciences Anglais Disponible
DisponibleEfficient Deep Self-Supervised Learning for Epileptic Seizures Detection Using EEG Signals / Zakarya Boudraf
Titre : Efficient Deep Self-Supervised Learning for Epileptic Seizures Detection Using EEG Signals Type de document : texte imprimé Auteurs : Zakarya Boudraf, Auteur ; Oussama houssem eddine Tamourt, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse Année de publication : 2023 Importance : 1 vol (48 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Self-supervised
epilepsyIndex. décimale : 004 Informatique Résumé : When dealing with epileptic seizures electroencephalogram’s (EEG) can
provide essential information about electrical brain activity. Patients with
epilepsy produce an excess of electrical discharges, causing involuntary body
movements along with other unpleasant symptoms. That information allows
us to detect these seizures and help minimize the damage done. However, an
EEG requires extensive knowledge of signal processing in order for medical
practitioners to correctly diagnose a patient. That’s why in recent years machine
learning algorithms have been employed to automatically detect these
seizures. But, in order to train these models, a large dataset of labeled data
is needed which is both expensive and time consuming. In this paper, we propose
three self-supervised learning models that make use of unlabeled data,
while still providing accurate and efficient predictions. Our model performs
competitively with previously published works by achieving an accuracy of
99.08%, sensitivity of 83.33%, and a FPR of 0.06.Côte titre : MAI/0700 En ligne : https://drive.google.com/file/d/1FUXzTw8ZQSN8HsU2J9S4jjfMtCBnM1b2/view?usp=drive [...] Format de la ressource électronique : Efficient Deep Self-Supervised Learning for Epileptic Seizures Detection Using EEG Signals [texte imprimé] / Zakarya Boudraf, Auteur ; Oussama houssem eddine Tamourt, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse . - 2023 . - 1 vol (48 f .) ; 29cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Self-supervised
epilepsyIndex. décimale : 004 Informatique Résumé : When dealing with epileptic seizures electroencephalogram’s (EEG) can
provide essential information about electrical brain activity. Patients with
epilepsy produce an excess of electrical discharges, causing involuntary body
movements along with other unpleasant symptoms. That information allows
us to detect these seizures and help minimize the damage done. However, an
EEG requires extensive knowledge of signal processing in order for medical
practitioners to correctly diagnose a patient. That’s why in recent years machine
learning algorithms have been employed to automatically detect these
seizures. But, in order to train these models, a large dataset of labeled data
is needed which is both expensive and time consuming. In this paper, we propose
three self-supervised learning models that make use of unlabeled data,
while still providing accurate and efficient predictions. Our model performs
competitively with previously published works by achieving an accuracy of
99.08%, sensitivity of 83.33%, and a FPR of 0.06.Côte titre : MAI/0700 En ligne : https://drive.google.com/file/d/1FUXzTw8ZQSN8HsU2J9S4jjfMtCBnM1b2/view?usp=drive [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0700 MAI/0700 Mémoire Bibliothéque des sciences Anglais Disponible
DisponibleFully Attention Convolutional Deep Neural Networks for Polyp Segmentation and Classification from Histological and Colonoscopic Images / Ines Mansour
Titre : Fully Attention Convolutional Deep Neural Networks for Polyp Segmentation and Classification from Histological and Colonoscopic Images Type de document : texte imprimé Auteurs : Ines Mansour, Auteur ; Nour Hamache, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse Année de publication : 2023 Importance : 1 vol (177 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Colorectal cancer
Colorectal polypsIndex. décimale : 004 Informatique Résumé : Colorectal cancer (CRC) is a type of cancer that affects the colon or rectum,
which are parts of the large intestine. It was the 3rd most commonly diagnosed
cancer worldwide and the 2nd leading cause of cancer-related deaths after lung
cancer in 2022. It can be prevented if glandular tissue (adenomatous polyps) is
detected early. At the same time, Colonoscopy has been strongly recommended
as a screening test for both early cancer and adenomatous polyps. However, its
limitations include the high polyp miss rate for smaller (<10mm) or flat polyps,
which are easily missed during visual inspection. Machine and Deep learning have
emerged as promising tools in the field of cancer research and diagnosis and can
be easily used in this context for the detection and segmentation of this type of
cancer.
In this work, we develop both Machine learning and Deep learning models that
can segment and classify medical images to improve the diagnosis and treatment
of colorectal cancer. We used Segmentation, and Classification techniques on
both Colonoscopic and Histological images. For that, we propose two deep CNN
convolutional neural network architectures from scratch,also we proposed CNN
model with and attention mechanism such as CNN with CBAM, and we tuned 12
pre-trained models in histological images.in totle we proposed 18 models for
histological images, and 6 others in colonoscopic images, that have been trained
on large datasets, such as:
→ Machine Learning Classifiers such as ANN, KNN, SVM, RF, ABD, and XGB.
→ CNN Architectures without Attention mechanism : CNN Models from
scratch.
→ CNN Architectures with Attention mechanism : CNN with CBAM.
→ Transfer Learning such as U-Net, DeepLabV3+, VGG, ResNet.
→ Transformers such as ViT (Vision Transformer): ViT B16, ViT B32.
→ Object Segmentation Methods such as SAM (Segment Anything Model).
iii
Our proposed models in Binary Classification on histological images obtained
accuracy in a range [84.93%, 99,80%]. And our proposed models in Semantic
Segmentation on colonoscopic images obtained an IoU score in the range [97.22%,
98.47%]. AI will revolutionize the medical field by enabling faster and more accurate
diagnoses, personalized treatment plans, and improved patient outcomes.Côte titre : MAI/0712 Fully Attention Convolutional Deep Neural Networks for Polyp Segmentation and Classification from Histological and Colonoscopic Images [texte imprimé] / Ines Mansour, Auteur ; Nour Hamache, Auteur ; Abdelouaheb Moussaoui, Directeur de thèse . - 2023 . - 1 vol (177 f .) ; 29cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Colorectal cancer
Colorectal polypsIndex. décimale : 004 Informatique Résumé : Colorectal cancer (CRC) is a type of cancer that affects the colon or rectum,
which are parts of the large intestine. It was the 3rd most commonly diagnosed
cancer worldwide and the 2nd leading cause of cancer-related deaths after lung
cancer in 2022. It can be prevented if glandular tissue (adenomatous polyps) is
detected early. At the same time, Colonoscopy has been strongly recommended
as a screening test for both early cancer and adenomatous polyps. However, its
limitations include the high polyp miss rate for smaller (<10mm) or flat polyps,
which are easily missed during visual inspection. Machine and Deep learning have
emerged as promising tools in the field of cancer research and diagnosis and can
be easily used in this context for the detection and segmentation of this type of
cancer.
In this work, we develop both Machine learning and Deep learning models that
can segment and classify medical images to improve the diagnosis and treatment
of colorectal cancer. We used Segmentation, and Classification techniques on
both Colonoscopic and Histological images. For that, we propose two deep CNN
convolutional neural network architectures from scratch,also we proposed CNN
model with and attention mechanism such as CNN with CBAM, and we tuned 12
pre-trained models in histological images.in totle we proposed 18 models for
histological images, and 6 others in colonoscopic images, that have been trained
on large datasets, such as:
→ Machine Learning Classifiers such as ANN, KNN, SVM, RF, ABD, and XGB.
→ CNN Architectures without Attention mechanism : CNN Models from
scratch.
→ CNN Architectures with Attention mechanism : CNN with CBAM.
→ Transfer Learning such as U-Net, DeepLabV3+, VGG, ResNet.
→ Transformers such as ViT (Vision Transformer): ViT B16, ViT B32.
→ Object Segmentation Methods such as SAM (Segment Anything Model).
iii
Our proposed models in Binary Classification on histological images obtained
accuracy in a range [84.93%, 99,80%]. And our proposed models in Semantic
Segmentation on colonoscopic images obtained an IoU score in the range [97.22%,
98.47%]. AI will revolutionize the medical field by enabling faster and more accurate
diagnoses, personalized treatment plans, and improved patient outcomes.Côte titre : MAI/0712 Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0712 MAI/0712 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible