Abstract: The present invention provides methods for universal lossy compression that provide performance at or near the rate-distortion limit and that are based on universal, implementable lossy source coding algorithms.
Type:
Grant
Filed:
February 5, 2010
Date of Patent:
November 27, 2012
Assignee:
The Board of Trustees of the Leland Stanford Junior University
Abstract: The present invention provides methods for universal lossy compression that provide performance at or near the rate-distortion limit and that are based on universal, implementable lossy source coding algorithms.
Type:
Application
Filed:
February 5, 2010
Publication date:
August 5, 2010
Applicant:
The Board of Trustees of the Leland Stanford Junior University
Abstract: Denoising such as discrete universal denoising (DUDE) that scans a noisy signal in an attempt to characterize probabilities of finding symbol values in a particular context in a clean signal can perform a rough denoising on the noisy signal and identify contexts from a roughly denoised signal. The rough denoising improves estimation of the statistical properties of the clean signal by reducing the false differentiation of contexts that noise can otherwise create. Statistical information regarding occurrences of symbols in the noisy signal and corresponding contexts in the roughly denoised signal can then be used to denoise the noisy signal. The specifics of the rough denoising can be chosen based on knowledge of the noise or of the clean data. Alternatively, the DUDE can be used in an iterative fashion where the denoised signal produced from a prior iteration provides the contexts for the next iteration.
Type:
Grant
Filed:
July 12, 2005
Date of Patent:
March 3, 2009
Assignee:
Hewlett-Packard Development Company, L.P.
Inventors:
Itschak Weissman, Erik Ordentlich, Gadiel Seroussi, Marcelo Weinberger, Sergio Verdu, Giovanni Motta