Nuance Communications, Inc.

International Business Machines Corporation

Tetsuya Uda

Takatoshi Watanabe

Noriko Imoto

Yoshinori Tahara

Daisuke Tomoda

Norikazu Ichikawa

A method and system for synchronizing words in an input text of a speech with a continuous recording of the speech. A received input text includes previously recorded content of the speech to be reproduced. A synthetic speech corresponding to the received input text is generated. Ratio data including a ratio between the respective pronunciation times of words included in the received text in the generated synthetic speech is computed. The ratio data is used to determine an association between erroneously recognized words of the received text and a time to reproduce each erroneously recognized word. The association is outputted in a recording medium and/or displayed on a display device.

Synchronization of an input text of a speech with a recording of the speech

A sounds-like spelling group is generated in which scores are used for word descriptions. A user selects a correct sounds-like spelling from a generated sounds-like spelling group, and a group of base forms is generated with scores provided for the selected sounds-like spelling and the reading of the word. In this fashion, a base form having a score that exceeds a reference value is registered without a voice recording being required. During the speech recognition process as performed by a user, when a predetermined error state is detected, e.g., when the number of recognition errors exceeds N, a panel is output requesting that the user register the inscription, the sounds-like spelling and the pronunciation inscription for the word and that the user record a corresponding pronunciation. In accordance with the pronunciation provided by the user, the base form is obtained and is re-registered in the speech recognition dictionary.

Methods and apparatus for recognized word registration in accordance with speech recognition

Growing community of inventors

Kanagawa, Japan

Tetsuya Uda