System and method for speech activated navigation

• US 20050131675A1
• Filed: 10/22/2004
• Published: 06/16/2005
• Est. Priority Date: 10/24/2001
• Status: Active Grant

• Current Assignee: Promptu Systems Corporation
• Original Assignee: Promptu Systems Corporation
• Inventors: Dubreuil, Jerome, Julia, Luc E., Bing, Jehan G.
• Granted Patent: US 7,289,960 B2
• Time in Patent Office: Days
• US Class Current: 704/200
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

The invention discloses a system and method for speech-activated navigating or browsing via a speech control interface used in a speech-activated multifunctional communications system. In one embodiment, the invention provides an approach to extend speech-activated navigation by linking an output of an open vocabulary recognizer to an Internet search engine in order that a user may have more options to search information related to his spoken commands. In another embodiment, the invention provides a means to enable the user to orally navigate a database via a speech control interface wherein the selections and associated selection criteria are organized into a hierarchical view menu. In another embodiment, the invention provides an approach with high flexibility and accuracy to recognize the user'"'"'s command using a new grammar structure and a matching score system.

Claims

1. A computer readable storage medium encoded with instructions, which when loaded into a digital computational device establishes a speech-activated Web browsing system, the system comprising:

a speech recognition system which is designed to recognize an input utterance against a list of candidate textual titles;

means for extracting textual titles from a given Web page, each of said textual title representing a content item which can be rendered on screen when a command associated with said textual title is activated;

means for generating a grammar for said textual titles extracted from said Web page;

means for applying said grammar to said speech recognition system;

wherein said speech recognition system processes the input utterance and determines its confidence level against a pre-set confidence level;

wherein if the confidence level of the input utterance is higher than said pre-set confidence level, then the textual title corresponding to the input utterance is recognized and the command associated with the textual title corresponding to the input utterance is activated;

wherein if the confidence level of the input utterance is lower than said pre-set confidence level, then said speech recognition system passes the input utterance through an open vocabulary recognizer, which is designed to transcribe arbitrary statements or texts that do not match any particular grammar path; and

means for taking an output of said open vocabulary recognizer as an input to an Internet search engine, said Internet search engine returning a Web page containing a number of hits of Internet link options.

2. A method for extending speech-activated Web browsing using a speech control interface, the method comprising the steps of:

extracting textual titles from a given Web page, each of said textual title representing a content item which can be rendered on screen when a command associated with said textual title is activated;

generating a grammar for said extracted textual titles, wherein each textual title is associated with a grammar path;

entering an input utterance to a speech recognition system which is designed to recognize the input utterance against a list of candidate textual titles;

applying said grammar to said speech recognition system;

processing the input utterance and determining its confidence level against a pre-set confidence level;

if the confidence level the input utterance is higher than said pre-set confidence level, then recognizing the textual title corresponding to the input utterance and activating the command associated with the textual title corresponding to the input utterance;

if the confidence level of the input utterance is lower than said pre-set confidence level, passing said input utterance through an open vocabulary recognizer, which is designed to transcribe arbitrary statements or texts that do not match any particular grammar path;

taking an output of said open vocabulary recognizer as an input to an Internet search engine; and

returning a Web page containing a number of hits of Internet link options.