Electrical Engineering and Systems Science > Image and Video Processing
[Submitted on 2 Feb 2022 (this version), latest version 11 Oct 2022 (v2)]
Title:Unpaired Image Super-Resolution with Optimal Transport Maps
View PDFAbstract:Real-world image super-resolution (SR) tasks often do not have paired datasets limiting the application of supervised techniques. As a result, the tasks are usually approached by unpaired techniques based on Generative Adversarial Networks (GANs) which yield complex training losses with several regularization terms such as content and identity losses. We theoretically investigate the optimization problems which arise in such models and find two surprising observations. First, the learned SR map is always an optimal transport (OT) map. Second, we empirically show that the learned map is biased, i.e., it may not actually transform the distribution of low-resolution images to high-resolution images. Inspired by these findings, we propose an algorithm for unpaired SR which learns an unbiased OT map for the perceptual transport cost. Unlike existing GAN-based alternatives, our algorithm has a simple optimization objective reducing the neccesity to perform complex hyperparameter selection and use additional regularizations. At the same time, it provides nearly state-of-the-art performance on the large-scale unpaired AIM-19 dataset.
Submission history
From: Milena Gazdieva [view email][v1] Wed, 2 Feb 2022 16:21:20 UTC (14,292 KB)
[v2] Tue, 11 Oct 2022 16:15:58 UTC (14,817 KB)
Current browse context:
eess.IV
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.