计算机科学
人工智能
图像去噪
翻译(生物学)
概率逻辑
扩散
图像(数学)
计算机视觉
单位(环理论)
统计物理学
数学
物理
生物
量子力学
遗传学
信使核糖核酸
基因
数学教育
作者
Hiroshi Sasaki,Chris G. Willcocks,Toby P. Breckon
出处
期刊:Cornell University - arXiv
日期:2021-01-01
被引量:2
标识
DOI:10.48550/arxiv.2104.05358
摘要
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 Fréchet Inception Distance (FID) performance on several public datasets, including both colour and multispectral imagery, significantly outperforming the contemporary adversarial image-to-image translation methods.
科研通智能强力驱动
Strongly Powered by AbleSci AI