Width-Wise Parameter Sharing for Multi-Domain GAN Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Width-Wise Parameter Sharing for Multi-Domain GAN Learning

Stratégie d'apprentissage multi-domaine pour les modèles génératifs antagonistes

Résumé

In this work, we propose a new parameter efficient sharing method for the training of GAN generators. While there has been recent progress in transfer learning for generative models with limited data, they are either limited to domains close to the original one, or adapt a large part of the parameters. This is somewhat redundant, as the goal of transfer learning should be to re-use old features. In this way, we propose width wise parameter sharing, which can learn a new domain with ten times fewer trainable parameters without a significant drop in quality. Previous approaches are less flexible than our method and also fail to preserve image quality for challenging transfers. Finally, as our goal is ultimately parameter re-use, we show that our method performs well in the multi-domain setting, wherein several domains are learned simultaneously with higher visual quality than the state of the art StarGAN-V2.
Fichier principal
Vignette du fichier
00_icipmain.pdf (786.41 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03983601 , version 1 (11-02-2023)

Identifiants

Citer

Ryan Webster, Julien Rabin, Loïc Simon, Frédéric Jurie. Width-Wise Parameter Sharing for Multi-Domain GAN Learning. 2022 IEEE International Conference on Image Processing (ICIP), Oct 2022, Bordeaux, France. pp.4163-4167, ⟨10.1109/ICIP46576.2022.9897423⟩. ⟨hal-03983601⟩
29 Consultations
54 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More