Research on International Standardization of Spoken Language Database and Assessment Techniques for Speech Input/Output

点击翻译按钮获取中文摘要

The aim of the project is to investigate the possibility of international coordination and eventual standardization of 1) spoken language databases that are being constructed in various countries, and 2) assessment techniques for speech recognition and synthesis.In the fiscal year 1990, after discussion with the Japanese members, Fujisaki visited all the foreign members in September for detailed discussion, In November all but one members gathered in Kobe for a symposium and workshop to discuss the state of the art, areas for possible standardization, and ways for international cooperation. In the fiscal year 1991, further visits were made by Fujisaki to U. S., Canada, Europe and China for detailed discussion. In September, all but two members gathered in Genova for a second workshop to discuss specific problems, methods, and mechanisms for international coordination. The followings are summaries of these series of discussions.1. International coordination and eventual standardization is obviously necessary, and have to be maintained by a permanent body consisting of official representatives from each of the concerned country or region.2. In order to establish such a permanent body, the following committees are tentatively formed.a) An advisory committee consisting of seven members (Fourcin, Fujisaki, Kurematsu, Liberman, Mariani, Millar, and Pallett) to select or recommend official representatives.b) Three working committees on speech database, speech input assessment, and speech output assessment, each consisting of several members, to work out detailed issues for international coordination and standardization.3. To pursue the aim of the project, efforts should be continued beyond the fiscal year 1992.