UNIT-DDPM: UNpaired Image Translation with Denoising Diffusion Probabilistic Models

TL;DR

This paper introduces UNpaired Image Translation with Denoising Diffusion Probabilistic Models (UNIT-DDPM), a stable, non-adversarial method for high-quality unpaired image translation that outperforms existing adversarial approaches.

Contribution

It presents a novel unpaired image translation framework using diffusion models without adversarial training, enabling stable training and superior image quality.

Findings

Achieves state-of-the-art FID scores on multiple datasets.

Outperforms adversarial methods in image quality and stability.

Works effectively on color and multispectral images.

Abstract

We propose a novel unpaired image-to-image translation method that uses denoising diffusion probabilistic models without requiring adversarial training. Our method, UNpaired Image Translation with Denoising Diffusion Probabilistic Models (UNIT-DDPM), trains a generative model to infer the joint distribution of images over both domains as a Markov chain by minimising a denoising score matching objective conditioned on the other domain. In particular, we update both domain translation models simultaneously, and we generate target domain images by a denoising Markov Chain Monte Carlo approach that is conditioned on the input source domain images, based on Langevin dynamics. Our approach provides stable model training for image-to-image translation and generates high-quality image outputs. This enables state-of-the-art Fr\'echet Inception Distance (FID) performance on several public datasets, including both colour and multispectral imagery, significantly outperforming the contemporary adversarial image-to-image translation methods.