JPS62247657A - Voice recognition and answering method - Google Patents

Abstract

PURPOSE:To confirm the operation of a voice recognition/answer telephone by having a manual abbreviated dial operation to synthesize the voices of words corresponding to the registered number of said abbreviated dial in case a voice recognizing function is not available. CONSTITUTION:When an abbreviated dial operation is carried out by means of a key 17a of a telephone set main body part 17 after a handset 16 is picked up, a transmission control part 11 sends the registered telephone number to a telephone line. At the same time, a synthesization indicating part 19 detects the operation of the abbreviated dial and gives a request to a voice synthesizing part 14 for voice synthesization of the words corresponding to the registered dial number. Then the part 14 reads a pattern out of a memory 15 and produces the signals of synthetic voices to send them to the part 11. The part 11 sends these signals to the handset 16 to give an answer in voices. Thus a correct operation of an abbreviated dial is confirmed.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a voice recognition response method suitable for use in telephones widely used in offices, homes, automobiles, and the like.

BACKGROUND OF THE INVENTION FIG. 2 shows the configuration of a conventional voice recognition answering telephone. 21 is a call control unit, 22 is a voice recognition unit, 23 is a memory for standard patterns for recognition, 24 is a voice synthesis unit, 25 is a memory for synthesis patterns, 26 is a telephone handset, and 27 is a normal Butsuphone telephone. Part corresponding to the main body, 28
is the registration control section.

Note that the call control unit 21 is further connected to the telephone line.

Next, the operation of the above conventional example will be explained. First, before recognizing and responding to a voice, the voice is registered.

At this time, the mode changeover switch 28 of the registration control unit 28
Set a to registration mode. Voice input is via the handset 26
from the voice recognition unit 22 through the registration control unit 28.
, and the speech synthesis unit 24, where it is converted into parameters used for each of recognition and synthesis, and then registered in the standard pattern memory 23 for recognition and the memory 25 for synthesis patterns. The voice to be registered is the name of the other party whose phone number is registered using speed dialing,
For example, they are "office", "home", etc., and are registered in correspondence with the speed dial registration numbers 01, 02, . . . . Registration of the telephone number of the other party corresponding to the speed dial registration numbers 01, 02, etc. is done in the telephone main unit 27 (2) as usual, but this telephone number is not registered at the central office but at the control station. Next, the operation when voice dialing is performed using voice recognition will be described.When the mode changeover switch of the registration control section 28 is set to the recognition mode, a recognition waiting state is entered.

Here, when voice is input from the handset 26, it is sent to the voice input unit 22 through the registration control unit 28, converted into parameters, and compared with the standard pattern of each word registered in the memory 23 to perform word recognition. Ru. The recognition result is sent to the call control unit 21, and based on the result (2), the call control unit 21 sends the telephone number registered with the registration number of the speed dial corresponding to the word to the telephone line.

The above recognition result is also sent to the speech synthesis section 24, and the speech synthesis section 2
4 reads the pattern from the memory 25 based on the result, creates a synthesized sound signal and sends it to the transmission control section 21, which sends it to the handset 26 for a voice response and receives the result of voice recognition. can be confirmed. If the recognition result is incorrect, by immediately putting the handset 16 on-hook, erroneous telephone connection can be prevented.

In this way, even with the conventional voice recognition response method described above, voice dialing and voice confirmation of the recognition result can be performed.

Problems to be Solved by the Invention However, with the conventional voice recognition response method described above, voice recognition may be difficult due to large ambient noise, or the voice recognition function cannot be used by a person other than the person who registered the voice. When a person tried to make a phone call, it was without any benefit.

The present invention solves these conventional problems,
The object of the present invention is to provide an excellent voice recognition response method that allows the voice response function to be used even when the voice recognition function cannot be used as described above.

Means for Solving the Problems In order to achieve the above object, the present invention is designed to respond by synthesizing a word corresponding to the registration number of the speed dial when a manual speed dial operation is performed. be.

According to the present invention, when it is not possible to use the voice recognition function, it is possible to perform a manual speed dial operation and confirm whether or not the operation was correct using a voice response. Since the same one is used for checking the recognition result, it has the effect that no extra steps are required in advance.

Embodiment FIG. 1 shows the configuration of an embodiment of the present invention. In the same figure, 11
12 is a call control unit, 12 is a voice recognition unit, 13 is a memory for standard patterns for recognition, 14 is a voice synthesis unit, 15 is a memory for synthesis patterns, 16 is a telephone handset, and 17 is a normal Butsuphone telephone body. 18 is a registration control section, and 19 is a composition instruction section. Note that the transmission control section 11
is further connected to a telephone line. Here, in the configuration of this embodiment, the difference from the conventional example is that a synthesis instruction section 19 receives a signal from a section 17 corresponding to the main body of a pushphone telephone and prompts a speech synthesis section 14 to perform synthesis. It is existence.

Next, the operation of the above embodiment will be explained. Since the voice registration, voice recognition, and confirmation of the recognition result by voice response are the same as the conventional example, there are some differences, that is, after manual speed dial operation. Let's talk about voice responses. It is assumed that the voice registration and the speed dial telephone number have already been registered, and the mode changeover switch 18a of the registration control section 18 is in the recognition mode. If you lift the transmitter/receiver 16 at this point, it will be in a state where voice recognition is possible, but if you perform a speed dial operation from the keys 17a of the telephone body 17 without inputting voice, the call controller 1
1 sends the telephone number registered with the registration number of the speed dial to the telephone line, but at the same time, the synthesis instruction unit 19 detects that the speed dial operation has been performed and sends the number of the word corresponding to the registration number of the speed dial. Speech synthesis unit 14 performs speech synthesis.
request.

This is similar to the synthesis request from the speech recognition section 12 to the speech synthesis section 14 when the speech recognition section 12 performs speech recognition. Based on this, the speech synthesis section 14 reads the pattern from the memory 15, creates a synthesized sound signal, and sends it to the transmission control section 11.The transmission control section 11 sends it to the handset 16 to perform a voice response, thereby shortening the signal. You can check the correctness of dial operations. If it is not correct, immediately put the handset 16 on-hook to prevent an erroneous telephone connection.

As described above, according to the above embodiment, in addition to being able to confirm voice dialing and its recognition results, when a speed dial operation is performed, it is possible to confirm that the operation is correct through a voice response. Moreover, this has the advantage of not requiring any extra effort.

In the above embodiment, the telephone number corresponding to the speed dial registration number is stored in the call control unit 11, but when connecting to a button button (PB) line,
The same applies if the central office memorizes the code and directly sends the shortened code.

Furthermore, in the above embodiment, the memory 13 for the standard recognition pattern and the memory 15 for the synthesis pattern are provided separately, but if the recognition parameters and synthesis parameters are the same, the memory 13 is provided separately. It can also be shared.

Effects of the Invention As is clear from the above-mentioned embodiments, the present invention not only allows voice dialing and confirmation of the voice recognition results, but also allows confirmation of correctness of speed dialing operations when the speed dialing operation is performed manually. , can be performed using a voice response similar to the voice response for confirming the recognition result, and furthermore,
Therefore, there is an advantage that no extra effort is required. This is an effective way to increase convenience when voice recognition is difficult due to high ambient noise, or for people other than the person who registered the voice, that is, people who cannot use the voice recognition function. be.

In addition, the present invention is particularly effective when using a car phone twice.This is because while driving a car, there is a lot of ambient noise depending on the conditions, making voice recognition difficult, and the memory of the speed dial number is vague. This is because it is difficult to confirm in this case.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a voice recognition response method according to an embodiment of the present invention, and FIG. 2 is a block diagram of a conventional voice recognition response method. 11... Call control unit, 12... Voice recognition unit, 13,
15...Memory, 14...Speech synthesis unit, 16...
Handset/receiver, 17... 1-speaker main unit, 17a... Key, 18... Registration control unit, 18a... Mode changeover switch, 19... Combination instruction unit. Agent's name: Patent attorney Toshio Nakao
1 is Figure 2

Claims (1)

[Claims] Word speech input from the handset is recognized and registered as a pattern for recognition and synthesis, and speed dial numbers for telephones input by voice are also stored in correspondence with the patterns, and the speed dial numbers from the handset are stored. A voice recognition response method, in which a voice response and a call are made to a voice input of a number, and a voice response and a call are made to a key input of the speed dial number using a synthesized voice according to the synthesis pattern.