Accès chercheur

EEDIS Laboratory

Evolutionary Engineering


Distributed Information Systems

Réseaux et Communication

Sécurité et Multimédia

Ingénierie des Connaissances

Data Mining & Web Intelligent

Interopérabilité des Systèmes d’information
& Bases de données

Développement Orienté Service

Deep multi-task learning for image/video distortions identification

Auteurs: » Ameur Zoubida
» Hamidouche Wassim
Type : Chapitre de Livre
Edition : Springer London ISBN:
Lien : »
Publié le : 26-10-2021

Identifyingdistortions in images and videos is important and useful in various visualapplications, such as image quality enhancement and assessment techniques.Instead of applying them blindly, these techniques can be applied or adjusteddepending on the type of distortion identified. In this paper, we propose adeep multi-task learning (MTL) model for identifying the types of distortion inboth images and videos, considering both single and multiple distortions. Theproposed MTL model is composed of one convolutional neural network (CNN) sharedbetween all tasks and N parallel classifiers, where each classifier isdedicated to identify a type of distortion. The proposed architecture alsoallows to adjust the number of tasks according to the number of distortiontypes considered, making the solution scalable. The proposed method has beenevaluated on natural scene images and laparoscopic videos â€¦

Tous droits réservés - © 2019 EEDIS Laboratory