Speech input communication system, user terminal and center system

• US 20040078202A1
• Filed: 02/14/2003
• Published: 04/22/2004
• Est. Priority Date: 06/20/2000
• Status: Active Grant

• Current Assignee: Sharp Kabushiki Kaisha (Hon Hai Precision Industry Co., Ltd.)
• Original Assignee: Sharp Kabushiki Kaisha (Hon Hai Precision Industry Co., Ltd.)
• Inventors: Kamiya, Shin
• Granted Patent: US 7,225,134 B2
• Time in Patent Office: Days
• US Class Current: 704/275
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

High recognition capability is obtained even with voice input from a user terminal low in power consumption. A user system (1) incorporating a voice instruction recognition processing portion (4) and a transmission control portion (6) is connected to a center system (2) incorporating a reception control portion (7), a voice instruction recognition processing portion (5), and an output control portion (8) through a communication line (3). Processing for obtaining a voice instruction content based on a waveform analysis result of the voice instruction is shared by voice instruction recognition processing portions (4, 5). This prevents decrease of a recognition rate like the case of disposing a voice analysis recognition processing function only on the center system (2) side. Also, there is eliminated waiting time for transfer of lexical grammar data generated in the case of disposing the voice analysis recognition processing function only on the user system (1) side, and a processing load of the user system (1) side is decreased. This makes it possible to accept a voice instruction with a large lexis even if the user system (1) is a mobile device such as cell phones that are low in power consumption.

Claims

1. A voice-input communication system having a user system with a user terminal and a center system being connected to the user system through a first communication line for sending an instruction from the user terminal to the center system, the voice-input communication system comprising:

voice instruction recognition processing means composed of a plurality of partial voice instruction recognition processing means having different functions for recognizing an instruction content from a result of waveform analysis of an inputted voice instruction, wherein

the plurality of the partial voice instruction recognition processing means are dispersed to the user system and the center system.

2. The voice-input communication system as defined in claim 1, wherein

the user system is provided with a user-side system connected to the user terminal through a second communication line, and

the user-side system is connected to the center system through the first communication line.

3. The voice-input communication system as defined in claim 1, comprising

a plurality of the center systems, each of which is connected through a third communication line.

4. The voice-input communication system as defined in claim 3, wherein

at least a final-stage center system is provided with output control means for outputting an instruction content recognized by the voice instruction recognition processing means.

5. The voice-input communication system as defined in claim 2, wherein

either the user terminal or the user-side system in the user system is provided with voice recognition means as the partial voice instruction recognition processing means for recognizing an inputted voice and outputting an interim recognition result, and provided with transmission control means for transmitting the interim recognition result to the center system through the first communication line.

6. The voice-input communication system as defined in claim 5, wherein

at least one center system comprises:

a lexical grammar memory for storing lexical grammar; and

language processing means as the partial voice instruction recognition processing means for performing language processing of the interim recognition result from the user system with use of the lexical grammar.

7. The voice-input communication system as defined in claim 5 or 6, wherein

the voice recognition means is phoneme recognition means, and

the interim recognition result is a phonemic lattice.

8. The voice-input communication system as defined in claim 7, wherein

as the interim recognition result, a phonemic similarity series per analytical frame or an average phonemic similarity series per set of sequenced multiple analytical frames is used instead of the phonemic lattice.

9. The voice-input communication system as defined in claim 5 or claim 6, wherein

the voice recognition means is syllable recognition means, and

the interim recognition result is a syllabic lattice.

10. The voice-input communication system as defined in claim 2 or claim 3, wherein

the first communication line is different in kind from the second and third communication lines.

11. The voice-input communication system as defined in claim 2, comprising:

phonation instruction means provided on at least one center system for creating a voice synthesis control parameter for a phonation content to be phonated by a user and transmitting the parameter to the user system through the first communication line;

voice synthesis means provided on the user terminal in the user system for creating a synthetic voice with the phonation content based on the voice synthesis control parameter transmitted through the first communication line;

waveform analysis means provided on either the user terminal or the user-side system in the user system for extracting characteristics of an inputted voice through analysis of a waveform of the inputted voice and transmitting the characteristics to the center system through the first communication line; and

speaker recognition means provided on the center system equipped with the phonation instruction means for recognizing a speaker based on the characteristics of the inputted voice transmitted through the first communication line.

12. A user terminal comprising:

phoneme recognition means for recognizing an inputted voice per phoneme and creating a phonemic lattice; and

transmission control means for transmitting the phonemic lattice through a communication line.

13. A center system for transmitting and receiving information to/from a user system through a communication line, comprising:

reception means for receiving a phonemic lattice transmitted through the communication line;

a lexical grammar memory for storing lexical grammar; and

language processing means for performing language processing of the phonemic lattice with use of the lexical grammar.