University Sétif 1 FERHAT ABBAS Faculty of Sciences
Détail de l'auteur
Auteur Abdallah Manallah |
Documents disponibles écrits par cet auteur
Ajouter le résultat dans votre panier Affiner la recherche
Plant Species Identification Using Siamese Network Architecture and Deep Boltzmann Machine Algorithm / Anes Manallah
Titre : Plant Species Identification Using Siamese Network Architecture and Deep Boltzmann Machine Algorithm Type de document : texte imprimé Auteurs : Anes Manallah, Auteur ; Abdallah Manallah ; Abdelouahab Moussaoui, Directeur de thèse Editeur : Sétif:UFS Année de publication : 2023 Importance : 1 vol (126 f.) Format : 29 cm Langues : Anglais (eng) Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 - Informatique Résumé : Deep learning techniques face challenges with small data sets. Recent proposals include
single-shot learning, under-learning, and Siamese networks. This paper introduces the
novel use of the Boltzmann mechanism (RBM) to explore efficient feature extraction techniques . A two-part Boltzmann machine (DBM) architecture is proposed for unsupervised
feature extraction,
Next, a trending technique called Siamese Neural Network is utilized, developing a
new Siamese technology based on Convolutional Neural Network (CNN). Training and
validation are conducted using plant datasets such as PlantVillage, Flavia, which measures
similarity among plant species based on leaf images. The CSN’s learned similarity function
is then generalized to classify both new and few-image species.
Plant recognition experiments are performed, training and validating CSN with datasets
from the Flavia dataset.Côte titre : MAI/0802
En ligne : https://drive.google.com/file/d/1lSxHjgCWOE6GI1G0h3u0yPHWhGenq0u4/view?usp=drive [...] Format de la ressource électronique : Plant Species Identification Using Siamese Network Architecture and Deep Boltzmann Machine Algorithm [texte imprimé] / Anes Manallah, Auteur ; Abdallah Manallah ; Abdelouahab Moussaoui, Directeur de thèse . - [S.l.] : Sétif:UFS, 2023 . - 1 vol (126 f.) ; 29 cm.
Langues : Anglais (eng)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 - Informatique Résumé : Deep learning techniques face challenges with small data sets. Recent proposals include
single-shot learning, under-learning, and Siamese networks. This paper introduces the
novel use of the Boltzmann mechanism (RBM) to explore efficient feature extraction techniques . A two-part Boltzmann machine (DBM) architecture is proposed for unsupervised
feature extraction,
Next, a trending technique called Siamese Neural Network is utilized, developing a
new Siamese technology based on Convolutional Neural Network (CNN). Training and
validation are conducted using plant datasets such as PlantVillage, Flavia, which measures
similarity among plant species based on leaf images. The CSN’s learned similarity function
is then generalized to classify both new and few-image species.
Plant recognition experiments are performed, training and validating CSN with datasets
from the Flavia dataset.Côte titre : MAI/0802
En ligne : https://drive.google.com/file/d/1lSxHjgCWOE6GI1G0h3u0yPHWhGenq0u4/view?usp=drive [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0802 MAI/0802 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible