Voice recognition rate increasing method in call initation of voice recognition calling device

2/04466

BACKGROUND OF THE INVENTION

1. Domain of 1¹ invention

The present invention relates to a voice recognition calling device and, more particularly, to a voice processing method for a call on a telephone of a voice recognition calling device comprising a telephone, so as to increase a voice recognition device. voice recognition rate in a call trigger.

2. Description of the technology concerned

In general, a calling device such as a hands-free telephone basically provides the functions of recognizing the voice received by a telephone, processing the voice and providing the voice output by the telephone. Therefore, this calling device necessarily comprises a telephone comprising a speaker and a microphone, a voice recognition processor for recognizing the voice received by the microphone, and a telephone processor for outputting the voice received by the microphone. the phone's microphone.

FIG. 1 is a block diagram of such a calling device, that is to say a device for recognizing the voice and for initiating a call. This voice recognition calling device basically recognizes voice words such as for example calling open, receiving and so on. The primary voice recognition calling device is a hands-free device. Another voice recognition calling device consists of an automatic telephone answering machine which will be marketed later.

A voice recognition paging device receives ringtones from a switch as described below. First, when ring tones are transmitted by the switch, the voice recognition paging device generates the ring tones as long as those ring tones are received.

Here, when a user enters a command such as voice calling, opening, or receiving, the voice recognition calling device recognizes the voice, switches

2/64466 in telephone mode and receive the ringtones. So the user can talk on the phone.

However, when the user wishes to speak on the phone, that user enters a voice command while the calling device is ringing. Thus, ringtone sounds and voice command are mixed. A general voice recognition paging device is very unlikely to pick up voice command in ringing and mixed voice command. This is due to the fact that the frequency of a ringtone generated in the general voice recognition paging device such as that shown in Figure 1, is typically a double frequency of 500 Hz and 1 kHz being in the same range. than voice, and that a one-second bell-cut follows a two-second bell-cut in its ringing cycle.

SUMMARY OF THE INVENTION

An object of the present invention is to create a voice processing method which can increase a voice recognition rate to enable ring tones to be received by a telephone in a voice recognition paging device.

Another object of the present invention is to create a voice processing method which makes it possible to initiate a call on a telephone even in the event of a fortuitous entry by a user of a voice command into a call recognition device. of voice.

To achieve the above objects, a voice processing method is provided for making a call on a telephone in a voice recognition calling device. In this method, a decrease in the recognition rate of the voice received for a call on a telephone is avoided by separating the frequency of a ringtone from that of a bandwidth of voice, or by shortening the ringtone.

To this end, the present invention relates to a voice processing method for a call on a telephone in a voice recognition call device comprising both a telephone, characterized in that it comprises the steps of:

- differentiate the frequency bandwidth of a ringing sound generated by a ringing and received, coming from a switch, in the voice recognition calling device, from the frequency bandwidth of the voice; and

- receive the ringtone in the telephone when entering a voice for a call on the telephone.

According to other characteristics of the invention:

- the frequency of the ringing sound is set to 1.8-2.2 kHz.

- the ringing tone is received in the telephone when a voice having a frequency of 300 Hz to 1.5 kHz is introduced for a call on the telephone.

The invention also relates to a voice processing method for a call on a telephone in a voice recognition call device comprising a telephone, characterized in that it comprises the steps of:

- generate a ringing with a cycle of 0.5 seconds of ringing on and 5.5 seconds of ringing cutoff, when a ringing is received, from a switch, in the voice recognition call device ; and

- receive the ringtone generated through the telephone when entering a voice for a call on the telephone.

According to another characteristic of the invention, this method comprises the steps consisting in:

- differentiate the frequency bandwidth of a ringing sound generated by a ringing and received, coming from a switch, in the voice recognition calling device, from the frequency bandwidth of the voice;

- generate a ringing with a cycle of 0.5 seconds of ringing on and 5.5 seconds of ringing cutoff; and

- receive the ringtone generated through the telephone when entering a voice for a call on the telephone.

2/04466

According to still other characteristics of the invention:

- the frequency of the ringing sound is set to 1.8-2.2 kHz.

- the generated ringtone is received through the telephone when entering a voice with a frequency of 300 Hz to

1.5 kHz for a phone call.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be described in more detail hereinafter with the aid of embodiments shown in the accompanying drawings in which:

- Figure 1 is a block diagram of a voice recognition calling device to which the present invention applies.

FIGS. 2A and 2B are views referred to to describe a method of processing input voice for a call on a telephone, according to an embodiment of the present invention; and

FIGS. 3A and 3B are views referred to in order to describe a method of processing input voice for a call on a telephone, according to another embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERENTIAL EMBODIMENTS

Preferred embodiments of the invention will now be described in detail with reference to the accompanying drawings. The detailed description of well-known functions and associated structures will not be repeated, when it is believed that this may obscure the object of the present invention. In addition, the terms used in the description are defined by considering the functions of the present invention. Thus, it will be appreciated that, as the terms may vary according to user, chip designer or common uses, the definitions should be based on the overall description of the present invention.

Figs. 2A and 2B are views referred to to describe a method of processing input voice in a call initiation for a call on a telephone, according to a first embodiment of the present invention.

2/6446«

This method is derived from the procedure whereby ring tones are generated by ring tone generator 22 of Figure 1, and transmitted to a voice recognition processor 14 through a speaker SPK and a microphone MIC of a telephone. In this method, the voice bandwidth filtered by a low-band filter 12 is improved to minimize the amount of voice introduced into a ringtone. Thus, the voice bandwidth filtered by the band pass filter 12 is set from 300 Hz to 1.5 kHz. A blocking frequency of the voice bandwidth filtered by the conventional bandpass filter 12 was 3.4 kHz. From the frequency analysis of the current voice, it can be seen that most of the voice energy is distributed in a range of 400-900 Hz, as shown in Figure 2A. As a result, frequency components of about 1.5 kHz or more of the voice frequency bandwidth can be blocked without any adverse results.

Further, in this method, the frequency of a ringtone is differentiated from that of a voice bandwidth, thereby blocking the ringtone reapplied through the MIC microphone, to prevent it from being fed into the recognition processor. of votes 14. Thus, the ringing frequency is made different from that of the voice bandwidth by using a double frequency of 1.8 kHz and 2.2 kHz exceeding the bandwidth range of a blocking frequency of the pass filter. -Band 12 when the ringing frequency resonates.

Thus, even if the ringtone is reapplied to the microphone MIC by the speaker SPK, this ringtone is output only by the processor 18 of the telephone and the speaker SPK, without being applied to the voice recognition processor 14 . In this way, even if there is a concurrent input of a voice intended for a call on a telephone, and a ringing coming through the microphone MIC, only the voice intended for the telephone call is applied to the recognition processor. voice 14. As a result, the controller 24 only receives the voice intended for the telephone call processed in the voice recognition processor 14, con2 / b44bb confirms the voice and controls the telephone processor 18 to transmit the ringtones to the telephone. Thus, a call is made possible in the telephone.

Figures 3A and 3B are views referred to to describe a method of processing input voice in a call initiation for a call on a telephone, according to another embodiment of the present invention.

In this method, even with a bell having a cycle of one second of bell on and two seconds of bell cutoff, as shown in Figure 3A, introduced by a switch, the controller 24 applies a bell having a cycle of 0 , 5 seconds of ringing on and 5.5 seconds of ringing cutoff, as shown in FIG. 3B. By generating a ringtone shorter than that which is output by the switch, i.e. by shortening the ringing cycle, it is possible to suppress a zone comprising a mixture of an input voice and a ringtone blocking the recognition of the input voice. As a result, a reduction, induced by the ringing, in the voice recognition rate can be avoided.

As described above, the present invention advantageously avoids decreasing the recognition rate of an input voice for a call on a telephone, by separating the frequency of a ringtone from that of a voice bandwidth, or by shortening the ringtone.