Image-to-image translation (I2I) aims at transferring the content representation from an input domain to an output one, bouncing along different target domains. Recent I2I generative models, which gain outstanding results in this task, comprise a set of diverse deep networks each with tens of million parameters. Moreover, images are usually three-dimensional being composed of RGB channels and common neural models do not take dimensions correlation into account, losing beneficial information. In this paper, we propose to leverage hypercomplex algebra properties to define lightweight I2I generative models capable of preserving pre-existing relations among image dimensions, thus exploiting additional input information. On manifold I2I benchmarks, we show how the proposed Quaternion StarGANv2 and parameterized hypercomplex StarGANv2 (PHStarGANv2) reduce parameters and storage memory amount while ensuring high domain translation performance and good image quality as measured by FID and LPIPS scores. Full code is available at https://github.com/ispamm/HI2I.
Hypercomplex image- to- image translation / Grassucci, Eleonora; Sigillo, Luigi; Uncini, Aurelio; Comminiello, Danilo. - (2022), pp. 1-8. (Intervento presentato al convegno 2022 International Joint Conference on Neural Networks, IJCNN 2022 tenutosi a Padua; Italy) [10.1109/ijcnn55064.2022.9892119].
Hypercomplex image- to- image translation
Eleonora Grassucci
;Luigi Sigillo;Aurelio Uncini;Danilo Comminiello
2022
Abstract
Image-to-image translation (I2I) aims at transferring the content representation from an input domain to an output one, bouncing along different target domains. Recent I2I generative models, which gain outstanding results in this task, comprise a set of diverse deep networks each with tens of million parameters. Moreover, images are usually three-dimensional being composed of RGB channels and common neural models do not take dimensions correlation into account, losing beneficial information. In this paper, we propose to leverage hypercomplex algebra properties to define lightweight I2I generative models capable of preserving pre-existing relations among image dimensions, thus exploiting additional input information. On manifold I2I benchmarks, we show how the proposed Quaternion StarGANv2 and parameterized hypercomplex StarGANv2 (PHStarGANv2) reduce parameters and storage memory amount while ensuring high domain translation performance and good image quality as measured by FID and LPIPS scores. Full code is available at https://github.com/ispamm/HI2I.File | Dimensione | Formato | |
---|---|---|---|
Grassucci_Hypercomplex_Pre-print_2022.pdf
accesso aperto
Tipologia:
Documento in Pre-print (manoscritto inviato all'editore, precedente alla peer review)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
6.35 MB
Formato
Adobe PDF
|
6.35 MB | Adobe PDF | |
Grassucci_Hypercomplex_2022.pdf
solo gestori archivio
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
6.35 MB
Formato
Adobe PDF
|
6.35 MB | Adobe PDF | Contatta l'autore |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.