Neural Network Compression; Deep Learning; Edge Devices
Résumé :
[en] Efficient model compression techniques are required to deploy deep neural networks (DNNs) on edge devices for task specific objectives. A variational autoencoder (VAE) framework is combined with a pruning criterion to investigate the impact of having the network learn disentangled representations on the pruning process for the classification task.
Centre de recherche :
Interdisciplinary Centre for Security, Reliability and Trust (SnT) > CVI² - Computer Vision Imaging & Machine Intelligence
Disciplines :
Physique, chimie, mathématiques & sciences de la terre: Multidisciplinaire, généralités & autres Ingénierie électrique & électronique Ingénierie, informatique & technologie: Multidisciplinaire, généralités & autres
Auteur, co-auteur :
SHNEIDER, Carl ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
ROSTAMI ABENDANSARI, Peyman ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
KACEM, Anis ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
SINHA, Nilotpal ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
SHABAYEK, Abd El Rahman ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
AOUADA, Djamila ; University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
Co-auteurs externes :
no
Langue du document :
Anglais
Titre :
Impact of Disentanglement on Pruning Neural Networks
Date de publication/diffusion :
20 juin 2023
Nombre de pages :
A0
Nom de la manifestation :
International Symposium on Computational Sensing (ISCS23)
Organisateur de la manifestation :
Thomas Feuillen, Amirafshar Moshtaghpour
Lieu de la manifestation :
Luxembourg, Luxembourg
Date de la manifestation :
12-06-2023 to 14-06-2023
Manifestation à portée :
International
Focus Area :
Computational Sciences Security, Reliability and Trust
