U.S. Philips Corporation

Lernout & Hauspie Speech Products N.v.

Manfred Grabherr

William F Ganong, Iii

Heribert Wutte

Brian Wilson

Ramesh Sarukkai

Gabor Janek

A modeless large vocabulary continuous speech recognition system is provided that represents an input utterance as a sequence of input vectors. The system includes a common library of acoustic model states for arrangement in sequences that form acoustic models. Each acoustic model is composed of a sequence of segment models and each segment model is composed of a sequence of model states. An input processor compares each vector in a sequence of input vectors to a set of model states in the common library to produce a match score for each model state in the set, reflecting the likelihood that a state is represented by a vector. The system also includes a plurality of recognition modules and associated recognition grammars. The recognition modules operate in parallel and use the match scores with the acoustic models to determine at least one recognition result in each of the recognition modules. The recognition modules includes a dictation module for producing at least one probable dictation recognition result, a select module for recognizing a portion of visually displayed text for processing with a command, and a command module for producing at least one probable command recognition result. An arbitrator uses an arbitration algorithm and a score ordered queue of recognition results, together with their associated recognition modules, to compare the recognition results of the recognition modules to select at least one system recognition result.

System and method for modeless large vocabulary speech recognition

In a product including a speech recognition device for controlling at least one processing unit while using a command lexicon containing commands which depends as regards its contents on an activated operating state of a processing unit and which speech recognition device includes a command lexicon memory in which commands of the command lexicon can be stored in the form of command information and identification information assigned to the latter, and which speech recognition device includes speech recognition means for detecting command information&mdash;contained in speech information representing a spoken command&mdash;and delivers command information to the at least one processing unit, generating means are provided which are arranged for generating&mdash;after an operating state of a processing unit has been activated&mdash;which may be effected with the aid of command information detected in a spoken command&mdash;a command lexicon assigned to only this one activated operating state and for storing this command lexicon in the command lexicon memory.

Product including a speech recognition device and method of generating a command lexicon for a speech recognition device

Growing community of inventors

Somerville, MA, United States of America

Manfred Grabherr