Systems and methods for speech recognition and separate dialect identification

• US 7,155,391 B2
• Filed: 05/24/2004
• Issued: 12/26/2006
• Est. Priority Date: 07/31/2000
• Status: Expired due to Term

• Current Assignee: Ovonyx Memory Technology, LLC (Micron Technology, Inc.)
• Original Assignee: Micron Technology, Inc.
• Inventors: Taylor, George W.
• Primary Examiner(s): Storm, Donald L.
• Application Number:
  US10/852,638
• Publication Number:
  US 20040215456A1
• Time in Patent Office: 946 Days
• Field of Search: 704/257, 704/235
• US Class Current: 704/257
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A speech-to-text conversion system. The two-way speech recognition and dialect system comprises a computer system, an attached microphone assembly, and speech-to-text conversion software. The two-way speech recognition and dialect system includes a database of dialectal characteristics and queries a user to determine their likely dialect. The system uses this determination to reduce the time for the system to reliably transcribe a user'"'"'s speech into text and to anticipate dialectal word usage. In another embodiment of the invention, the two-way speech recognition and dialect system is capable of transcribing the speech of multiple speakers while distinguishing between the different speakers and identifying the text belonging to each speaker.

Claims

1. A system for transcribing spoken words to text, the system comprising:

at least one data structure containing word data correlating words to speech signals, wherein the at least one data structure includes dialect parameter data that is different than the word data and that can be used to recognize user dialects; and

a processor configured to receive speech signals and to produce text representations corresponding to spoken words, wherein the processor initially sends signals to a user asking the user at least one question indicative of the user'"'"'s dialect to thereby induce the user to provide non-verbal answer signals to the processor that are indicative of dialect parameters, wherein the dialect parameters comprise at least one of an age of the user, a gender of the user, an educational level of the user, a native language of the user, a geographic origin of the user, and a current geographic residence of the user,

and wherein the processor compares the dialect parameters from the user to the dialect parameter data in the at least one data structure to determine at least one dialect that the user is likely to have prior to comparing the speech signals to the word data.

2. The system of claim 1, wherein the at least one data structure includes phonemic characteristic data.

3. A speech recognition system comprising:

a means for storing dialect data;

a means for processing speech signals into text representations, wherein the processing means receives at least one dialect parameter indicative of at least one of a user'"'"'s age, gender, educational level, native language, geographic origin, and current geographic residence,

wherein the processing means compares the at least one dialect parameter to the dialect data to determine at least one dialect that the user is likely to have, and wherein the processing means further compares speech signals to word data to produce text representations corresponding to spoken words of the user, said comparison of the speech signals following said determination of the at least one dialect; and

a means for obtaining input from the user, wherein the processing means receives the at least one dialect parameter via the obtaining means, and wherein the means for obtaining input comprises a keyboard.

4. A speech recognition system comprising:

a means for storing dialect data;

a means for processing speech signals into text representations, wherein the processing means receives at least one dialect parameter indicative of at least one of a user'"'"'s age, gender, educational level, native language, geographic origin, and current geographic residence,

wherein the processing means compares the at least one dialect parameter to the dialect data to determine at least one dialect that the user is likely to have, and wherein the processing means further compares speech signals to word data to produce text representations corresponding to spoken words of the user, said comparison of the speech signals following said determination of the at least one dialect; and

an interface means for sending signals to the user, wherein the processing means sends a first signal to the user via the interface means to thereby induce the user to provide a non-verbal answer signal that is indicative of the at least one dialect parameter.

5. A method of processing speech for a speech recognition system, the method comprising:

determining if one of a plurality of dialects has been assigned to a user;

if one of the plurality of dialects has not been assigned to the user;

prior to receiving speech signals from the user, receiving at least one dialect parameter from the user;

comparing the at least one dialect parameter to dialect parameter data to determine at least one of the plurality of dialects that the user is likely to have;

receiving speech signals corresponding to spoken words of the user; and

using the at least one of the plurality of dialects that the user is likely to have to produce text representations corresponding to the spoken words of the user.

6. The method of claim 5, wherein said receiving at least one dialect parameter further comprises receiving a non-verbal answer signal from the user.

7. The method of claim 6, wherein said receiving the at least one dialect parameter is performed through a computer peripheral device.

8. The method of claim 7, wherein the at least one dialect parameter is received through manual input by the user.

9. The method of claim 6, additionally comprising prompting the user to provide the non-verbal answer signal.

10. The method of claim 9, wherein said prompting comprises providing the user with a series of parameter queries.

11. The method of claim 10, wherein the series of parameter queries comprises at least three questions.

12. The method of claim 6, wherein the at least one dialect parameter is indicative of at least one of the user'"'"'s age, gender, educational level, native language, geographic origin, and current geographic residence.