Yissum Announces a Novel Method for Swift and Robust Pattern Recognition

Jerusalem, Israel – Yissum, the Technology Transfer Company of the Hebrew University of Jerusalem, introduced today a novel method for real time, automatic, generic and robust pattern matching. The novel algorithm, developed by Professor Michael Werman and Ofir Pele, both from the School of Computer Science and Engineering at the Hebrew University of Jerusalem, enables very rapid recognition of a particular pattern in a fraction of the time currently available. The novel algorithm can be used in computer vision software for managing images, in robotics as a simple and fast method for vision-based systems used for assembly manufacturing and inspection, as well as for face recognition and additional security applications. The findings were published in the journal IEEE Transactions on Pattern Analysis and Machine Intelligence (Pele, O. and Werman, M., August 2008, 30(8) 1427-1443).

"This novel method enables ultra-rapid pattern recognition which is highly robust and reliable," said Nava Swersky Sofer CEO of Yissum. "The algorithm can enhance various imaging and computer vision applications that are becoming an ever-growing part of everyday life. For example, it can be helpful in quick information retrieval from large visual databases. One such application can be photographing a restaurant and

immediately accessing relevant reviews. In the filed of security, the algorithm can be used, among others, for surveillance purposes by finding a suspected person in a video movie."

Many applications in image processing and computer vision require finding a particular pattern in an image, a process termed pattern matching. Pattern matching is typically performed by scanning the entire image, and evaluating a distance measure between the sought pattern and areas, or windows, in the image. The novel algorithm is much faster than current methods because it does not attempt to estimate the distances for non-similar windows, but only decides that these windows are non-similar. The reduction in running time is due to the fact that unnecessary information is not computed. The method is applicable to any pattern shape, even a non-contiguous one, and is automatic and robust, enabling detection of low quality patterns, rotated patterns or patterns that are partly occluded.

About Yissum

Yissum was founded in 1964 to protect the Hebrew University’s intellectual property and commercialize it. $1 Billion in annual sales are generated by products based on Hebrew University technologies licensed out by Yissum. Ranked among the top technology transfer companies in the world, Yissum has registered 5500 patents covering 1600 inventions; licensed out 480 technologies and spun out 65 companies. Yissum’s business partners span the globe and include companies such as Novartis, Microsoft, Johnson & Johnson, Merck, Intel, Teva and many more. 

For more information, please contact:

Tamir Huberman

www.yissum.co.il

Yissum Introduces a Novel Method for Ranking Online Reviews

JERUSALEM (Business Wire)

Suppose you want to go to a movie, or buy a new book, or perhaps a new digital camera. What would be your first step? Certainly, most of us would consult the internet for reviews about the specific product we want to buy. This is where things usually get complicated, because we are often swamped with hundreds or thousands of reviews, some of which are useless.

Now, Yissum Research Development Company Ltd., the technology transfer arm of the Hebrew University of Jerusalem, presents a new technology especially suitable for ranking product reviews according to the particular needs of the person performing the search. This novel method for content analysis, invented by Professor Ari Rappoport and Mr. Oren Tsur from the Hebrew University's School of Computer Science and Engineering, is specifically designed to automatically rank book and product reviews according to their estimated helpfulness.

"This novel method has the potential of changing the way we extract information from the internet. It already allows us to easily identify the most helpful reviews on various topics, and, in the future, will actually generate one comprehensive, optimized review from the available data. It offers a necessary tool for coping with the vast amount of data in the web world," said Yaacov Michlin, CEO of Yissum. "This is one out of dozens of breakthrough technologies developed at the Hebrew University and available for licensing in the field of computer science."

The technology, RevRank, first identifies a core of dominant terms that defines a virtual optimal review, and then uses those terms to rank other reviews relative to the `virtual core` review. The simplicity of RevRank enables easy understanding of the output and an effortless parameter configuration to match the personal preferences of different users. RevRank was successfully tested on book and product reviews on Amazon. As a result effective reviews that had a good chance of being missed by readers were moved up to top positions.

About Yissum

Yissum Research Development Company of the Hebrew University of Jerusalem Ltd. was founded in 1964 to protect and commercialize the Hebrew University`s intellectual property. Ranked among the top technology transfer companies in the world, Yissum has registered over 6,100 patents covering 1,750 inventions; has licensed out 480 technologies and has spun-off 65 companies including, BriefCam, HumanEyes, Mobileye and ReadEasy. Yissum`s business partners span the globe and include companies such as IBM, Intel, Johnson & Johnson, Merck, Monsanto, Novartis, Phillips, Roche, Sygenta, Teva Pharmaceuticals, Vilmorin and many more. For further information please visit www.yissum.co.il.

Yissum Ltd.

Tamir Huberman

Copyright Business Wire 2009