Continuous Methods : Hamiltonian Domain Translation
Not on the map. Why?
Abstract (arXiv)
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.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 9 of its 16 references that are in the map. First posted on arXiv 8 Jul 2022.
Builds on
9 of 16 references.
References · 9
- 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
- 2018Neural Ordinary Differential Equationscited by 3,779
- 2014Deep Learning Face Attributes in the Wildcited by 3,693
- 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,614
- 2015Variational Inference with Normalizing Flowscited by 2,170
- 2018Spectral Normalization for Generative Adversarial Networkscited by 1,636
- 2019Hamiltonian Neural Networkscited by 712
- 2018Invertible Residual Networkscited by 365
- 2020Scalable Gradients for Stochastic Differential Equationscited by 271
Built on by
Citing papers per year: 2023: 1.
1 paper.