Spotlight
in
Workshop: Continuous Time Perspectives in Machine Learning
Continuous Methods : Hamiltonian Domain Translation
Emmanuel Menier · Michele Alessandro Bucci · Mouadh Yagoubi · Lionel Mathelin · Marc Schoenauer
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.
Chat is not available.