Apprentissage des réseaux de neurones profonds sur des données multimodales
Type doc. :
Langue :
Auteur(s) :
Année de soutenance:
Thème :
Afficher le Résumé
Our perception of the world is most of the time multi-modal. In other words,natural intelligence is not limited to just a single modality. Our brain is able to process multiple types of information relevant to a specific task simultaneously,and the same applies to artificial intelligence. Overall, the integration of multi-modality holds great promise for advancing the capabilities of artificial intelligence in processing multi-modal information. This thesis discusses particularly the mechanisms of multi-modality in the context of Artificial Intelligence i.e. such as merging multi-modal information from different sensor modalities and points to the importance of multi-modal perception. Different Deep Learning models are trained and evaluated using data provided by the national electricity and gas company SONELGAZ. The performance of the various contributions is evaluated by comparing the empirical results of the different deep architectures.
Télécharger Document
| N° Bulletin | Date / Année de parution | Titre N° Spécial | Sommaire |
|---|
| Cote | Localisation | Type de Support | Type de Prêt | Statut | Date de Restitution Prévue | Réservation |
|---|---|---|---|---|---|---|
| 620.381 CHE TH C1 | BIB-Centrale / Thèses | Electronique | externe | disponible |