Recognition dictionary system structure and changeover method of speech

Abstract

A speech recognition system realizing large-vocabulary speech recognition at a low cost without deteriorating the rate of recognition and a recognition speed performance is provided with a dictionary change-over section for making a change-over between dictionaries to be subjected to speech recognition in accordance with dictionary change-over information, a first memory for storing a plurality of dictionaries, a second memory for storing one dictionary made an object of recognition, and a speech recognition section for performing a speech recognition processing, whereby speech recognition is performed while making a change-over between dictionaries, as required. For example, in a car navigation speech recognition system, the change-over between dictionaries is made for each area in accordance with position information.