JPH0536499U - Voice recognizer - Google Patents

Abstract

(57) [Summary] [Purpose] To allow a plurality of speakers to control one voice recognition device without automatically opening the individual memory of the registered speaker by automatically recognizing the voice of the speaker. Is possible. [Structure] The system control unit 7 stores a keyword pattern in a keyword memory unit 8. The individual word pattern is stored in the individual memory unit 9. In the voice recognition mode, first, the keyword of the speaker is input, the pattern is created by the pattern creating section 5, and the pattern recognition section 6 compares it with the pattern of each keyword stored in the keyword memory section 8 and inputs it. The speaker is determined. The system control unit 7 opens only the individual memory of the determined speaker, and recognizes the voice input after that based on the speaker's word stored in the memory.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention relates to a specific speaker voice recognition device, and more particularly to a voice recognition device that can be used by a plurality of speakers.

[0002]

[Prior Art]

 Conventionally, the specific-speaker voice recognition device can operate only one speaker who registered a voice pattern. In particular, when the voice recognition device is applied to car audio and the like, the use of the voice recognition device is limited to only registered persons such as drivers, which is inconvenient.

[0003]

[Problems to be solved by the device]

 The present invention has been made in view of the above points, and its purpose is to register a voice pattern and recognize a voice based on the pattern. It is to provide a voice recognition device that enables a person to operate the device.

[0004]

[Means for Solving the Problems]

 The speech recognition device of the present invention comprises a keyword memory section for storing the keyword pattern of each speaker to be registered and an individual memory section for storing the respective word pattern of each speaker. By inputting the pattern, the registered speaker is determined by comparing the pattern with the pattern in the keyword memory section, the individual memory section of the speaker is opened, and the pattern stored in the individual memory section is input. It recognizes voice.

Further, each speaker is provided with an individual memory unit that stores a pattern of words of each speaker to be registered, and a keyword pattern of each speaker is stored in a specific address of the individual memory unit. By comparing the pattern with the pattern at the specific address by inputting, the registered speaker is determined, the individual memory of the speaker is opened, and the voice input by the pattern stored in the individual memory is input. Is to recognize.

[0006]

[Action]

 According to the voice recognition device of the present invention, the keyword pattern of each speaker to be registered is stored in a specific address of the keyword memory unit or the individual memory unit, and each speaker inputs a key word to register the speaker. Is recognized by the keyword pattern, and the input voice is recognized by referring to the individual memory part of the speaker, so that the operation of multiple speakers is possible.

[0007]

【Example】

 An embodiment of this invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an on-vehicle audio equipment using a voice recognition device according to a first embodiment of the invention. The voice signal detected by the microphone 1 is amplified by the preamplifier 2, converted into a digital signal by the AD converter 3, and input to the voice feature extraction unit 4.

The voice feature extraction unit 4 extracts a feature from the input voice digital signal and outputs it to the pattern generation unit 5. The pattern creating section 5 creates an individual pattern from the characteristics of the audio digital signal and outputs it to the system control section 7.

The system control unit 7 inputs a signal from the pattern generation unit 5, and also receives signals from the keyword memory unit 8, individual memory units 9, 10, 11 and the pattern recognition unit 6 and the control signal generation unit 12. Input and output signals to control the entire voice recognition system.

The car audio device 13 is controlled by a signal from the control signal generation unit 12, reproduces a recording signal of a recording medium such as a CD, and sounds a speaker 14.

In the above configuration, the voice of each speaker is first registered. For example, when performing voice registration of Mr. A, B, and C, Mr. A inputs his initials as a keyword, and the pattern creation unit 5 outputs the pattern to the system control unit 7. The system control section 7 stores the keyword pattern in the keyword memory section 8. Mr. A further inputs individual words for controlling the car audio device, for example, a tape, a CD, a tuner, etc. by voice, and the pattern is stored in the individual memory unit 9.

When the voice registration of Mr. B and C is performed, the keyword pattern is stored in the keyword memory unit 8 and the individual word pattern is stored in the individual memory units 10 and 11, respectively.

In the voice recognition mode, first, a keyword of the speaker is input, the pattern is created by the pattern creating section 5, and the pattern of each keyword stored in the keyword memory section 8 is read by the pattern recognizing section 6. The input speaker is determined by comparison.

The system control unit 7 opens only the individual memory of the determined speaker, recognizes the voice input thereafter by the speaker's word stored in the memory, and the control signal generation unit 12 The car audio device 13 is controlled by the output signal.

FIG. 2 is a block diagram showing a configuration of an on-vehicle audio equipment using a voice recognition device according to a second embodiment of the invention. In this embodiment, the keyword pattern of each speaker is stored in the heads 9a, 10a, 11a of the individual memory section. Other configurations and operations are similar to those of the first embodiment. In this embodiment, when a speaker's keyword is input, the pattern creating section 5 creates the pattern, and the pattern recognizing section 6 creates a pattern for each keyword stored in the heads 9a, 10a, 11a of the individual memory sections. The input speaker is determined by comparing with the pattern. The system control unit 7 keeps only the individual memory of the determined speaker open and closes the other individual memories, and according to the speaker's word stored in the opened individual memory, the voice input thereafter is changed. The car audio device 13 is recognized and controlled by the signal output from the control signal generator 12.

[0016]

[Effect of the device]

 According to the voice recognition device of the present invention, a plurality of speakers can control one voice recognition device. In addition, the speaker's voice is automatically recognized and the registered speaker's individual memory is automatically opened, so there is no need to switch it with a switch or the like.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a vehicle-mounted audio device using a voice recognition device according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of an in-vehicle audio device using a voice recognition device according to a second embodiment of the present invention.

[Explanation of symbols]

 1 Microphone 2 Preamplifier 3 AD Converter 4 Voice Feature Extraction Section 5 Pattern Creation Section 6 Pattern Recognition Section 7 System Control Section 8 Keyword Memory Section 9 Individual Memory Section 12 Control Signal Generation Section 13 Car Audio Device 14 Speaker

Claims (2)

[Scope of utility model registration request] 1. A keyword memory unit for storing a keyword pattern of each speaker to be registered, and a respective individual memory unit for storing a word pattern of each speaker.
First, by inputting a keyword, the pattern is compared with the pattern of the keyword memory section to determine the registered speaker and open the individual memory section of the speaker,
A voice recognition device, which recognizes a voice input by a pattern stored in an individual memory unit. 2. An individual memory section for storing the word pattern of each speaker to be registered is stored, and the keyword pattern of each speaker is stored in a specific address of the individual memory section, and the keyword is input first. By comparing the pattern with the pattern of the specific address, the registered speaker is determined, the individual memory section of the speaker is opened, and the input voice is recognized by the pattern stored in the individual memory section. A voice recognition device characterized by: