JPH0361958B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a voice recognition device that can reliably recognize voice signals and pushphone signals input through a telephone line.

[Technical background of the invention and its problems]

BACKGROUND ART Recently, with the development of the information society, it has been considered to perform data processing by recognizing voice signals and pushphone signals emitted from telephones in systems connected via telephone lines. However, it is extremely difficult to recognize and process such signals, which have clearly different characteristics, such as a voice signal and a pushphone signal, according to a single algorithm.
Furthermore, improvement in recognition accuracy cannot be expected. Therefore, conventionally, as shown in Fig. 1, a pushphone signal recognition section 1 equipped with a dedicated algorithm for recognizing pushphone signals, a voice recognition section 2 equipped with a dedicated algorithm for recognizing voice signals, and a A speech recognition device is constituted by a comprehensive judgment section 3 that comprehensively judges the recognition results obtained by the recognition sections 1 and 2.

In this way, the number of words to be recognized by the speech recognition unit 2 can be arranged, so that recognition processing can be performed with a certain degree of reliability. However, if, for example, an ambiguous signal is input,
It is difficult for the pushphone signal recognition unit 1 to reliably reject this as a voice signal, and it is also difficult for the voice signal recognition unit 2 to similarly recognize this as a pushphone signal and reliably reject it. Since it is difficult to determine whether the input signal is a pushphone signal or an audio signal, the overall determination section 3 has a problem in that it cannot reliably identify whether the input signal is a pushphone signal or an audio signal. Furthermore, in order to eliminate such problems, it is necessary to improve the reject ability of each of the recognition sections 1 and 2, resulting in a problem that the device configuration becomes complicated. There was also the problem that the effects commensurate with such increased complexity were not expected to be that great.

[Purpose of the invention]

The present invention has been made in consideration of these circumstances, and its purpose is to provide a highly practical voice recognition device that can easily and reliably distinguish between voice signals and pushphone signals. There is a particular thing.

[Summary of the invention]

The present invention provides an audio signal recognition means equipped with an algorithm dedicated to recognizing audio signals and a pushphone signal recognition means equipped with an algorithm dedicated to recognizing pushphone signals. A determination means for determining whether or not it is a pushphone signal is provided, and a predetermined process is performed on the recognition result by the pushphone signal recognition means or the recognition result by the voice signal recognition means in accordance with the determination result by the determination means. It is.

〔Effect of the invention〕

Therefore, according to the present invention, the input signal is identified by utilizing the unique feature of the pushphone signal that the tone is stable in a predetermined time axis frame, and then the audio signal and the pushphone signal are respectively differentiated according to the identification results. Since recognition processing is performed separately,
The recognition accuracy is extremely high. Moreover, since the processing format is simple and the device configuration is simple, the reliability can be easily improved.
The practical advantages are enormous.

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

FIG. 2 is a schematic configuration diagram of the embodiment device, and the first
Components that are the same as those of the conventional device shown in the figure are designated by the same reference numerals. The feature of this embodiment device is that the determination unit 4 processes the input signal by calculating the similarity using the acoustic feature dictionary of the pushphone signal.
Based on this, it is determined whether the input signal is a pushphone signal or not.
The above similarity calculation process is performed on the input signal every predetermined time axis frame. The determination based on the similarity value is performed, for example, as a binary value of an audio signal/pushphone signal, or as a ternary value obtained by adding an undeterminable signal to the binary value. Such determination results are sent to the recognition units 1 and 2 and the comprehensive determination unit 3.

When the judgment result is that the pushphone signal is identified, the pushphone signal recognition section 1 is driven by the judgment signal to recognize the input signal. Then, the recognition result is outputted via the comprehensive determination section 3. Further, when the determination result is identified as a voice signal, the voice recognition unit 2 is driven by the determination signal. As a result, the input signal is voice recognized, and the recognition result is outputted via the comprehensive determination section 3. If an undeterminable determination result is obtained, the recognition units 1 and 2 are each driven, and a recognition result is determined in each of them. At this time, the comprehensive judgment section 3 performs a comprehensive judgment on both of the above recognition results according to a predetermined algorithm, and outputs the judgment result as the final recognition result for the input signal.

Thus, with the device configured as described above, it is possible to easily and accurately distinguish between an audio signal and a pushphone signal, which have clearly different properties, and then recognize the signal according to an appropriate algorithm for each case. can. Therefore, by dividing the recognition processing process for voice signals and pushphone signals, which has conventionally been extremely complicated, into different systems, it is possible to simplify the recognition process and improve the recognition accuracy. In other words, it is possible to easily improve the performance of the device.

By the way, the determination section 4 that identifies input signals as described above can be configured as shown in FIG. 3, for example. That is, the input signal is led to the preprocessing unit 11, and the input signal is analyzed every predetermined time axis frame of, for example, several tens of milliseconds, to obtain, for example, the bandpass filter output A and the low-frequency or full-band filter output B. . The bandpass filter output A is led to the similarity calculation section 12, and similarity calculation is performed with feature data for each category of the pushphone signal stored in the pushphone signal/acoustic signal dictionary 13.

On the other hand, the analysis section determining section 14 uses the full-band filter output B to determine an analysis processing section, for example, by detecting sections where the signal level is high. A control signal is given to the counting processing section 15 at the analysis start point and analysis end point. This counting processing unit 15 counts the number of times the similarity calculation unit 12 obtains a similarity value equal to or greater than a predetermined value θs within the interval set as described above. This determination of similarity values exceeding the predetermined value θs is performed for all categories. The counted number of times k is then given to the recognition determining section 16 together with the information l of the analysis section.

The recognition determination unit 16 has two thresholds θks(i) and θkw(i) depending on the value of the analysis interval l, and compares these thresholds with the count value k to determine the input signal. I'm going. However, i takes a value of 1, 2, 3...l. Then, we make three judgments: kθks(l)…(1) θks(l)＞kθkw(l)…(2) θkw(l)＞k…(3) If the above condition becomes (1), then has determined that the input signal is a pushphone signal. In addition, if the above condition (2) is satisfied, it is impossible to judge the input signal, and if the condition (3) is satisfied, the input signal cannot be determined.
In each case, a determination result is obtained that the input signal is an audio signal.

According to the determination results obtained in this manner, the recognition and determination processes described above are performed in the recognition units 1 and 2 and the comprehensive determination unit 3, respectively.

As described above, according to the present device, it is no longer necessary to provide a dictionary for pushphone signals in the speech recognition unit 2, which is the center of recognition processing, so that the processing speed is greatly improved compared to conventional devices, and the dictionary It is possible to significantly improve the recognition rate by increasing the degree of separation, and to simplify the device configuration by reducing the dictionary storage area. Further, even if the voice recognition section 2 has a low reject ability for pushphone signals, the pushphone signal recognition section 1 can easily perform the processing without considering the reject ability for voice signals. In other words, compared to audio signals with large feature variations, only pushphone signals with small feature variations can be processed, so that highly reliable pushphone signal recognition can be achieved using an extremely simple configuration. Furthermore, the configuration of the determining section 4 can be easily realized as shown in FIG. In addition, it is also possible to carry out the analysis interval determination process by directly using the preprocessing results of the device, that is, without providing a special preprocessing section 11 or the like as the determination section 4, which simplifies the configuration of the device as a whole. can be achieved. Therefore, it is possible to simplify dictionary processing and other related processing, and to improve the processing speed, which brings about great practical effects.

Note that the present invention is not limited to the above embodiments. For example, it goes without saying that the pushphone signal includes a pseudo pushphone signal using an acoustic coupler or the like. Furthermore, instead of having a dictionary for pushphone signals for all categories, it is also effective to mutually cluster the categories and provide a small number of them as a dictionary. Furthermore, if the degree of separation of the features between the audio signal and the pushphone signal is relatively large, the above-described processing may be performed while maintaining the predetermined identification performance. In this way, the time required for matching with the dictionary can be shortened, and the storage area for the dictionary can also be reduced, among other advantages.
As described above, the present invention can be implemented with various modifications without departing from the gist thereof.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing an example of a conventional device, FIG. 2 is a schematic configuration diagram of an embodiment of the device of the present invention, and FIG. 3 is a configuration diagram of a determining section in the embodiment device. DESCRIPTION OF SYMBOLS 1... Pushphone signal recognition section, 2... Audio signal recognition section, 3... Comprehensive judgment section, 4... Judgment section, 1
1... Preprocessing section, 12... Similarity calculation section, 13...
...Feature dictionary, 14... Analysis interval determination section, 15...
Feature section counting section, 16... Recognition determining section.

Claims (1)

[Claims] 1. In a voice recognition device connected to a telephone line into which voice signals and pushphone signals are input, a voice signal recognition means equipped with an algorithm dedicated to recognizing voice signals, and a pushphone equipped with an algorithm dedicated to recognizing pushphone signals. It is equipped with a signal recognition means and a determination means for determining whether or not a signal inputted through the telephone line is a pushphone signal by treating only the pushphone signal as a recognition processing target, and according to the determination result of the determination means, the voice A speech recognition device characterized in that it processes a recognition result of a signal recognition means or a pushphone signal recognition means.