A sampled texture prior for image super-resolution

Conference item

Super-resolution aims to produce a high-resolution image from a set of one or more low-resolution images by recovering or inventing plausible high-frequency image content. Typical approaches try to reconstruct a high-resolution image using the sub-pixel displacements of several low- resolution images, usually regularized by a generic smoothness prior over the high-resolution image space. Other methods use training data to learn low-to-high-resolution matches, and have been highly successful even in the single-input-image case. Here we present a domain-specific im- age prior in the form of a p.d.f. based upon sampled images, and show that for certain types of super-resolution problems, this sample-based prior gives a significant improvement over other common multiple-image super-resolution techniques.

Authors: Pickup, LC; Roberts, SJ; Zisserman, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: MIT Press
• Host title: Advances in Neural Information Processing Systems 16 – Proceedings of the 2003 Conference
• Pages: 1587-1594
• Place of publication: Cambridge, Massachusetts, USA
• Publication date: 2004-08-03
• Event title: Neural Information Processing Systems: Natural and Syththetic (NIPS 2003)
• Event location: Vancouver, Whistler, Canada
• Event website: https://neurips.cc/Conferences/PastConferences
• Event start date: 2003-12-08
• Event end date: 2003-12-13
• ISSN: 1049-5258
• ISBN-10: 0262201526
• ISBN-13: 9780262201520
• Language: English