KR0170529B1 - Voice-recognition automatic dialing system - Google Patents

Abstract

The voice recognition automatic dialing system is used in the mobile phone to input the phone number as a multi-syllable voice signal in Korean, English, Japanese or foreign language, and the recognized voice signals are sequentially stored in the buffer. With the first-in-first-out function, the voice recognition auto dialing system is improved compared to the portable telephone having the voice recognition auto dialing system which can only decode the sequence and then transmit it through the DTMF transmitter. Provides a portable telephone.

Description

Voice recognition automatic dialing system

1 is a block diagram of a dialing system using a conventional keypad.

2 is a block diagram of a voice recognition automatic dialing system according to the present invention.

Figure 3 is a block diagram showing in detail the configuration of the voice recognition automatic dialing system according to the present invention.

4 is a diagram illustrating a procedure in which call set-up is performed between a subscriber, a speech recognition system and an exchange using a speech recognition automatic dialing system in which a speech pattern of a single syllable isolated word according to the present invention is performed. .

5 is a diagram illustrating a procedure in which a call set-up is performed between a subscriber and a voice recognition system and an exchange using a voice recognition automatic dialing system in which a voice pattern of a compound tone is recognized according to the present invention.

* Explanation of symbols for main parts of the drawings

1: Mode selector switch 2: Keypad

3: speaker voice recognition system 4: DTMF transmitter

5: DTMF transmitter including decoder

6: base band 7: radio frequency (RF) stage

8: microphone 9: comparator

10: Amplifier (PRE-AMP) 11: Codec

12: ROM 13: Dual-Port SRAM

14: CPU (CPU) 15: Buffer (BUFFER)

16: voice synthesizer 17: speaker

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice recognition automatic dialing system of a portable telephone, and more particularly, to a voice recognition automatic dialing system that enables dialing by voice without visually checking a key pad.

In general, when using a mobile phone using a manual dialing system using a keypad while moving, there is a problem that the user has to check the number plate with the naked eye, thereby causing various accidents.

In the conventional voice recognition automatic dialing system designed to overcome the above problems, it is not necessary to check the number plate directly with the naked eye when moving, but the user's voice information should be registered in advance in the memory device of the voice recognition automatic dialing system, and the user himself or herself. Since the dialing is made only through the voice of the user, the user must directly input the information into the memory device, and only the user can make a call, and there is a problem that the number of memorable registration numbers is quite limited.

In order to solve the above problems, the voice recognition automatic dialing system according to the present invention uses a baseband and a radio frequency (RF) stage of a portable telephone system using a conventional keypad as it is, and uses a phone number recognized through the voice recognition system. After the input is completed in the buffer sequentially, after the input is completed using a first-in first-out function to sequentially decode the value to be input, and to send the decoded value through the DMP transmitter a voice recognition automatic dialing system It aims to provide.

In the present invention, in order to achieve the above object, a voice pattern of single syllable isolated words corresponding to 1,2 ‥‥ 9. 9 required for ROM dialing of a voice recognition automatic dialing system (one, two, three, ...), And voice patterns such as sound ends indicating completion of voice recognition dialing are input.

In addition, when an error occurs in the ordinal-odd composite speech speech recognition pattern (one, two, two, nine, nine) required for dialing into the ROM. Voice patterns, such as a note again for instructing redialing, are input, each of which has a voice pattern of pronunciation for English, Japanese, and other foreign languages, as well as Korean. It consists of a complex number of Ordinal Numbers, and the order is that ordinal-oddity and radix-ordinal order can be used, and voice recognition is performed by programming voice patterns of English, Japanese and other foreign languages as well as Korean. The number of possible languages can be increased according to the needs of the user.

In addition, a push-pull switch or a wireless remote control switch having a toggle switching function may be used as a specific implementation method of the mode switching switch, and the voice recognition automatic dialing system of the present invention may be installed in a vehicle. In this case, the driver inside the vehicle is installed in a position that is easy to control, and when the user touches or presses the mode switching switch, the dialing device is switched to the voice mode to perform the voice recognition auto dialing function, and the mode switching switch is reset again. When dialing or pressing, manual dialing function is performed through the keypad and DTMF transmitter so that the moving user can dial the desired phone number easily after touching the mode switching switch, so that they are not disturbed at all. Characterized in that the dialing can be performed.

The present invention is expressed in the negative phonetic patterns of English, Japanese, Japanese, Japanese, Korean, Japanese, Korean, Japanese, Japanese, Japanese, Korean, Japanese, Japanese, Korean, Japanese, Korean, Japanese, Korean, Korean, Chinese, Japanese, Korean, Chinese, Korean, Chinese, Korean, Korean, Japanese, Korean, Japanese, Korean, Korean In addition, the voice recognition sound is output through the voice synthesizer, and the central processing unit (CPU) disables the codec while the voice is output, thereby preventing the sound output through the speaker from being input into the microphone again.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 shows a dialing system using a conventional keypad employed in a portable telephone, showing a keypad 2, a DTMF transmitter 4, a base band 6, and a radio frequency (RF) stage. .

2 is a diagram illustrating a dialing system having a voice recognition function of a portable telephone according to the present invention, wherein a baseband 6 and a radio frequency (RF) stage 7 of a portable telephone using a conventional keypad are used as they are. 2) and a DTMF transmitter 5 having a speaker speech recognition system 3 and a decoder so as to be compatible with the DTMF transmitter 4, and to the mode switching switch 1 It shows up to DTMF transmitter which can be encoded / decoded by one module. The complex base band 6 and radio frequency (RF) stage 7 used in the portable telephone using the conventional keypad are left as it is. And, by configuring the recognition module as a DTMF transmitter 5 having a speaker voice recognition system 3 and a decoder, a user can arbitrarily select a keypad or a voice by means of a mode switch. And the module is connected to the output of the DTMF transmitter 4.

Therefore, when the user presses the mode switch, the voice recognition automatic dialing system of the present invention is switched to the voice mode to perform the voice recognition auto dialing function, and when the user presses the mode switch again, the keypad and the DTM F. Manual dialing function is performed through.

FIG. 3 shows the configuration of the speaker voice recognition system 3 of the voice recognition auto dialing system shown in FIG. 2 in detail. The voice input through the external microphone 8 of the voice recognition auto dialing system (MIC IN) is performed. The signal is amplified by the amplifier 10 (Pre-AMP) through the comparator 9, and passes through the codec 11, CODEC, the codec 11 is a band pass filter (BPF) having a bandwidth of 100HZ to 4KHZ The analog-to-digital converter (A / D) converter which converts the analog signal into a digital signal through the C1) and the digitally converted signal have a high correlation with the reference pattern stored in the ROM 12 through a pattern matching algorithm. It consists of a DSP controller that scans and retrieves patterns.

The matched value is stored in the Dual-Port SRAM 13 and the central processing unit 14 (e.g., INTEL8051) reads the matched value stored in the SRAM 13, SRAM. 16, for example, echo through AMPCM 4BIT, the user checks whether the recognition sound output through the speaker 17 is correctly recognized, and if the recognition is good, say the next number.

At this time, when the voice synthesizer 16 outputs the sound recognized through the voice output device, the central processing unit 14 disables the codec 11 so that the sound output through the voice output device is input to the microphone 17 again. Prevent it.

In addition, when the next input is input through the microphone, the CPU 14 checks whether the input value is equal to or greater than the reference value VREF of the comparator 9.

Here, the input signal of the reference value or more is determined to be a noise signal.

At this time, when a value equal to or greater than the reference value is input, it is determined that recognition is good. If the recognition is wrong, the user presses the recognize button on the keypad 2 (not shown) or inputs a voice signal called again, and inputs a phone number voice signal again.

On the other hand, the well-recognized voice data is stored in the buffer 15, and when the end of the input of all phone numbers of the telephone number is recognized, when the end sound is recognized, a Nm (Nom-maskable Interrupt) is applied to about 0.1. With a delay time of 2 seconds, the decoding value of the telephone number stored in the buffer 15 is transmitted toward the DTMF transmitter 5, and the transmitted signal is transmitted to the baseband 6 and the radio frequency (RF) stage. After passing through (7), the recognition sounds stored in the buffer 15 are dialed in order to enable a telephone call.

4 shows a call set-up procedure of a voice recognition automatic dialing system in which a speech pattern of a single syllable isolated word according to the present invention is recognized. (102) The user keys in the voice dial mode (103), and the voice recognition system instructs and waits for voice input (104). At this time, when the user utters the first digit sound, which is one of twelve isolated words of zero, one, two, nine, ball, and end (105), the speech recognition system recognizes, confirms, and responds (106). Next, the user checks the response voice (107), and if correctly recognized, the user pronounces the next number. The recognition system, on the other hand, considers the digits stored in the buffer as correct recognition if there is a next input and waits for the next input (108).

Then, the user utters the next digit (109), and the voice recognition system recognizes and confirms the digit as well and responds (110). At this time, when the user checks the digits and thinks that the recognition is wrong, when the user presses the Recognition button on the keypad (111), the voice recognition system does not store the recognition sounds in the buffer and waits for the next input. (SP, Stack Pointer) points to the next address recognized (112). Thereafter, in the voice recognition system, after the voice recognition instruction and waiting (113), the above-described process is repeated.

Finally, when the user recognizes all the digits, the user utters the end (114). When the end is recognized, the voice recognition system processes the NMI and sends out the numbers stored in the buffer one by one by delaying 0.1 seconds ( 115).

FIG. 5 is a flowchart illustrating a call set-up process of a voice recognition automatic dialing system in which another ordinal-odd or odd-odd composite voice pattern is recognized according to the present invention. Call set-up of the voice recognition automatic dialing system in which the voice pattern of a word is recognized is the same, but in the voice recognition automatic dialing system in which the voice pattern of a compound voice is recognized, the user inputs the air input by air, work, or two. ‥‥ 11 digits of the last word or foreign language words, and when the user checks the digits, if the user thinks that the recognition is wrong, pressing the recognize button on the keypad or inputting a voice signal called again (111) Only difference is that. For reference, Table 1 below shows an example of Korean and English among voice patterns programmed in the ROM 12.

By using the voice recognition automatic dialing system of the portable telephone according to the present invention as described above, it is possible to simply dial by voice without having to check the number plate with the naked eye.

The invention is particularly well suited for use in cellular phones which are now widely used. Recently, in the United States, it is prohibited by law to lift or dial a cell phone while driving a car. Because, when the driver dials while driving, the driver should look at the dial number and the front alternately, so it is very dangerous when driving at high speed, and the risk of accident is increased.

Therefore, when the voice recognition automatic dialing system according to the present invention is applied to a mobile phone attached to a vehicle, an accident can be prevented in advance.

In addition, the present invention is applied to the car navigation system (CNS) and the audio / video (A / V) device of the car to guide the route through the existing voice, the driver needs to press the touch panel or function keys one by one. You can create a useful system that can be easily manipulated without

In still another embodiment of the present invention, the speaker voice recognition device of the present invention can be applied to a TV, a fan, an air conditioner, or the like to operate a TV, a fan, an air conditioner, or the like as a voice, and the detailed description thereof is as follows.

That is, by introducing the speaker speech recognition device of the present invention into a TV employing a digitalized TV channel selection system or the like, it is possible to select a channel as voice.

In addition, by applying the speaker voice recognition system of the present invention to the air conditioner, the switching operation of the rotary switch can be operated by voice, so that the wind direction speed such as stop, strong wind, weak cooling, strong cooling, etc. can be selected, The speaker voice recognition system can be applied to a fan. The speed control winding can be used to adjust the wind speed and rotation of the fan by using a speed control switch to operate the voice of a speed change switch that adjusts the wind speed by a resistance size.

The present invention is also not limited to only the above-described embodiments, but may be variously applied to all electronic circuits in which switching operations are required by slightly changing or adding circuits and software, which are conventional in the art. Anyone with knowledge of this can easily know. Therefore, the present invention is not limited to the claims described below, all electronic circuits using the speaker voice recognition device may be said to belong to the scope of the present invention.

Claims (8)

A mode switching switch for recognizing the voice of the speaker and selecting one of the push button system mode and the voice recognition automatic dialing system mode, and when the push button system is selected by the mode switching switch, the voice signal is converted into several signals. DTMF transmitter for changing and transmitting, and DTMF transmitter with decoder for changing and transmitting voice signals into a plurality of decodable signals when the speaker voice recognition system mode is selected by the mode switching switch. A baseband for receiving signals from the DTMF transmitter and the decodable DTMF transmitter and filtering the baseband, and a radio frequency for changing the signal from the baseband; Voice recognition automatic dialing system including an RF stage. The digital speaker of claim 1, wherein the speaker speech recognition system comprises an amplifier for amplifying a voice signal input through a microphone, a bandpass filter for passing the amplified signal, and an analog signal which is a signal passing through the bandpass field. Analog-to-Digital Converter (A / D) for converting into a signal, and DSP (DSP) controller for scanning and searching for a pattern having a high correlation with the reference pattern stored in the ROM through a pattern matching algorithm A codec, a dual-port SRAM storing the value retrieved by the codec, and a value stored in the dual-top SRAM are read and echoed through a speech synthesizer. A speech synthesizer, a central processing unit (CPU) for re-entering the signal by pressing the resend button on the keypad if the echoed sound is not recognized, and all DTMF transmitter that transmits the decoded value of the telephone number stored in the buffer when the output is terminated, and a base that receives the decoding value transmitted from the DTMF transmitter and converts it into a digital field signal form. And a radio frequency (RF) stage for propagating and dialing the digital signal of the base band in the air at a radio frequency. The method of claim 2, wherein the automatic voice dialing system comprises: an input means for inputting a voice signal as a voice pattern of a single syllable isolated word corresponding to a telephone number and performing analog-to-digital conversion; and a number of 0 to 9 single syllable isolated words. Compares the rom stored in advance as a speech pattern of the speech pattern of the single syllable isolated word inputted through the input means with the speech pattern of the single syllable isolated word prestored in the ROM to obtain information of the highest correlation. Voice recognition automatic dialing system comprising a RAM for temporary storage. The voice recognition automatic dialing system according to claim 2, wherein the voice recognition automatic dialing system comprises input means for analog-to-digital conversion by inputting a voice signal corresponding to a telephone number as an ordinal-odd or radix-odd composite sound, and numbers 0 to 9. The highest correlation is obtained by comparing the ROM stored in the ordinal-odd or radix- ordinal composite pattern in advance with the speech pattern of the composite sound inputted through the input means and the composite pattern previously stored in the ROM. Voice recognition auto dialing system comprising a RAM for temporarily storing the information of the. The voice recognition dialing system of claim 3, further comprising a voice recognition completion sound indicating that the voice recognition completion sound of the telephone number to be input is completed. 5. The speech recognition completion method of claim 4, wherein the ROM recognizes that the voice recognition is wrong other than the numbers 0 to 9, and again, that the voice recognition is completed again. Voice recognition automatic dialing system characterized in that it further stores the voice pattern of the sound. The voice recognition automatic dialing system according to claim 4, wherein the re-recognition indication tone again and the end of the voice recognition completion tone can be stored in a foreign language corresponding thereto. The voice recognition automatic dialing system as claimed in claim 4, wherein the digits corresponding to the numbers 0 to 9 include Korean, English, Japanese, or all other languages.