![[Previous Section]](../icns/leftarrw.gif){kind=small}
![[Top of Report]](../icns/tocbttn.gif){kind=small}
![[Send Your Comments]](../icns/cmmntbt.gif){kind=small}
![[Welcome Page]](../icns/jtecbttn.gif){kind=small}
![[Next Section]](../icns/rghtarrw.gif){kind=small}
Site: Omron
Information Technology Research Center
Shimokaiinji, Nagaokakyo-City
Kyoto, 617, Japan
http://www.wg.omron.co.jp/index.html
Date Visited: 26 March 1998
WTEC Attendess: B. Croft (report author), R. Reddy, B. Davis-Brown, G. Mendel, R. Chellappa, R. Larsen
Hosts:
Omron is an international, diverse company with about $6 billion in annual sales and more than 22,000 employees. The main products are in the areas of control components and systems (more than 50% of sales), electronic fund transfer systems, public information and traffic systems, medical and healthcare equipment, and open systems. Products related to digital information organization technology would come under the last category. The four core technologies originally identified by the company are fuzzy technology, sensing technology, computers, communication and control technology, and life science technology.
Recently, "humanmedia" or "flexible intelligence" technology to support "human-oriented computing" has been identified as a new direction. This technology is being developed both to support other Omron products, such as ticket sales machines, and as stand-alone software products. The goal of Omron's humanmedia technology is to create "human-oriented equipment" to achieve the "Optimization Society," where people can enjoy work and private lives suited to their individual needs. Although digital libraries are of potential interest to the Multi Media Group within the information technology research center, Omron currently has no products in that area or plans to develop such products. Some Omron software products, however, are important components of current commercial retrieval systems in the Asian markets.
In terms of government funding, Omron receives funding from NEDO, which provided ¥22 million for the FY 1998 and was expected to provide ¥24 million during FY 1999.
Omron's "humanmedia" technology is a combination of language processing technology, speech recognition technology, image processing technology, and "kansei processing technology." Kansei is a term that the WTEC team encountered frequently during this visit and is related to the development of intuitive interfaces and to the concept of affective computing proposed by the MIT Media Laboratory. Each of these technologies is summarized in the following sections.
The topics that were discussed here include morphological analysis, multilingual input and language recognition. SuperMorpho-J is a Japanese morphological analysis system that is used for word segmentation and part-of-speech tagging. The system has high segmentation accuracy (about 98%) and can process more than 1 GB of text per hour. The system can also deal with some issues specific to Japanese such as word variants and word breaks at the end of the line. Versions of this system have been developed for Korean and Chinese. Supermorpho-J is being used in the Japanese versions of Infoseek and Verity's search engine. Although segmentation and different character encodings are often described as significant problems for Asian languages, it was clear from this discussion that these issues have been largely dealt with and are not an obstacle to multilingual retrieval systems.
Wnn is a well-known input system for Japanese and Chinese that was originally developed in the UNIX environment. This system has been steadily developed as a product and ported to Windows and Java. The system now also supports multiple languages in one document and many encodings, including Unicode. Once again, from the point of view of multilingual digital library systems, it is clear that input will not be a major issue.
A speech recognition project based on word spotting was described. Word spotting was preferred to large vocabulary language model approaches due to the potential for better robustness in noisy environments, such as information kiosks or ticket machines. Omron is now focusing its technology on Japanese language, and a detailed explanation was not provided.
An image retrieval project was described as Omron's part of a national project supported by NEDO. More than 13 companies, laboratories, and universities are involved in 3 working groups. The group that Omron was involved in is "kansei agent and human media database." The image retrieval project is viewed as a means of developing a better understanding of kansei rather than being leading edge research in image processing. The definition of kansei used by this group is "subjective criterion in human information selecting." Images from a database of postage stamps are indexed using four types of composition templates: horizontal, vertical, circle, and radiation. Combinations of these templates are being used for retrieval experiments, and excellent results have been obtained for the domain (postage stamps).