Continuous Methods : Hamiltonian Domain Translation - Laboratoire Interdisciplinaire des Sciences du Numérique Access content directly
Conference Papers Year : 2022

Continuous Methods : Hamiltonian Domain Translation

Abstract

This paper proposes a novel approach to domain translation. Leveraging established parallels between generative models and dynamical systems, we propose a reformulation of the Cycle-GAN architecture. By embedding our model with a Hamiltonian structure, we obtain a continuous, expressive and most importantly invertible generative model for domain translation.
Fichier principal
Vignette du fichier
example_paper.pdf (1.76 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03716629 , version 1 (07-07-2022)

Identifiers

  • HAL Id : hal-03716629 , version 1

Cite

Emmanuel Menier, Michele Alessandro Bucci, Mouadh Yagoubi, Lionel Mathelin, Marc Schoenauer. Continuous Methods : Hamiltonian Domain Translation. Workshop « Continuous time methods for machine learning », ICML, Jul 2022, Baltimore, United States. ⟨hal-03716629⟩
146 View
40 Download

Share

Gmail Facebook X LinkedIn More