JP2744039B2 - Voice recognition device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Industrial application field The present invention relates to a voice recognition device for a specific speaker provided with registration means having excellent operability.

(B) Conventional technology In a specific-speaker speech recognition device that a user registers and uses after recognition in advance, it is necessary for the user to input characters or numbers corresponding to the recognition words as recognition results in advance. But,
Since the input voice pattern and the recognition result display data are a set of associated data, a method has been adopted in which, for each word, the display data is input by a key following the voice registration of the recognition word. However, in this method, it is necessary to alternately utter a voice and input a key, which makes the operation cumbersome and increases the burden on the user.

(C) Problems to be Solved by the Invention Only registration of speech patterns of recognition words is performed at once,
After that, if the method of inputting the recognition result display data collectively and key input is used, the voice input and the key input can be performed separately, making the operation easy to understand, but when the number of words increases, which word is uttered first It is difficult to remember what happened, which is the problem.

(D) Means for Solving the Problems In the speech recognition device according to the present invention, when a recognition word is registered,
At the same time as analyzing and storing as a voice pattern for recognition,
Means for storing data analyzed as a voice pattern for synthesis; means for synthesizing the voice pattern for synthesis; and storage of display data of a recognition result input after voice synthesis in a position corresponding to the voice pattern for recognition. After the registration of the recognition words is performed collectively, the speech pattern for synthesis is synthesized in order from the beginning to input the display data of the recognition result corresponding to the recognition word. It becomes possible.

(E) Function According to the present invention, it is not necessary to remember the order of what recognition word was registered at the time of voice registration, and the voice input and the key input of the display data of the recognition result are performed separately. , It is possible to perform the registration at a time, so that the troublesomeness at the time of registration is reduced.

(F) Embodiment FIG. 1 shows an embodiment of the speech recognition apparatus of the present invention.

As shown in the figure, the flow of processing when voice registration of a recognized word and key input of recognition result display data are performed collectively at the time of voice registration will be described below.

The voice input from the microphone 11 is amplified by the amplifier 12 to such an extent that the amplitude is not saturated, and the voice analysis unit 13 for recognition is used.
Is sent to the speech analysis unit 14 for synthesis, and is analyzed into a recognition parameter and a synthesis parameter, respectively. Then, the recognition parameters are stored in the standard voice pattern memory 15,
The synthesis parameters are stored in the synthesis pattern memory 17. This process is repeated until registration of all recognized words is completed.

Next, synthesis patterns are sent to the speech synthesis unit 20 in order from the top of the synthesis pattern memory 17, and the recognition word is synthesized and output. After confirming the synthesized sound, the user inputs recognition result display data from the key input unit 18, and the input data is stored in the memory 18a as recognition result display data. Thus, the user checks the synthesized pattern synthesized and output by the voice synthesis unit 20 on the synthesized output display unit 19 while checking the synthesized pattern as a synthesized sound.

As described above, this processing is repeated until the input of the recognition result display data for all the recognized words is completed.

At the time of recognition, the voice input from the microphone 11 is amplified by the amplifier 12 to such an extent that the amplitude is not saturated, and is analyzed by the voice analysis unit 13 for recognition to create an input voice pattern. The matching unit 16 performs pattern matching between the input voice pattern and the standard voice pattern in the standard voice pattern memory 15, calculates a standard voice pattern with the shortest distance, and determines a recognition word. Then, the display unit 19 displays recognition result display data for the recognized word.

(G) Effect of the Invention As described above, according to the present invention, the voice pattern input of the recognition word at the time of registration and the key input of the recognition result display data can be performed at the same time. The annoyance is reduced.

In addition, by confirming the synthesized sound, it is possible to confirm that the input voice is correctly registered, so that the update of the registered voice can be omitted, and the burden on the user can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of a speech recognition apparatus according to the present invention. 11 ... Microphone, 16 ... Pattern matching part, 12
…… Amplifier, 17… Speech pattern memory for synthesis, 13… Speech analysis unit for recognition, 18… Data input unit for display, 14… Speech analysis unit for synthesis, 19 …… Display unit, 15 …… Standard Voice pattern memory, 20 ... Voice synthesis unit.

Continuing from the front page (72) Inventor Shinichi Tsuruto 2-18-18 Keihanhondori, Moriguchi-shi, Osaka Sanyo Electric Co., Ltd. (72) Inventor Mitsumi Okura 2-18-18 Keihanhondori, Moriguchi-shi, Osaka (56) References JP-A-63-292196 (JP, A) JP-A-59-147243 (JP, U)

Claims (1)

(57) [Claims] In a specific speaker voice recognition apparatus provided with voice input means and voice analysis means, a memory for storing a voice pattern for recognition at the time of voice registration, a voice for synthesizing an input voice corresponding to each voice pattern. Memory means for storing data, and memory means for storing display data for recognition results, and means for analyzing input speech into a speech pattern for recognition at the time of speech registration and simultaneously analyzing and storing the speech data for synthesis, Means for performing voice synthesis on the stored voice data for synthesis;
A means is provided to store the display data of the recognition result input after speech synthesis in a position corresponding to the recognition voice pattern, and the input voice is recorded when the recognition word is registered, and registration of all the recognition words is completed. A speech recognition apparatus characterized by inputting display data of a recognition result while reproducing in order from the beginning at a point in time.