JP2005148199A - Information processing apparatus, image forming apparatus, program, and storage medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve usability by making it easy to input a retrieval keyword when a file group is retrieved. <P>SOLUTION: When the retrieval keyword is inputted in voice (steps S3 and S4), the voice of the inputted retrieval keyword is recognized by a voice recognition part and converted into text data (step S6), which are displayed on a console panel (step S7). When it is considered that the displayed retrieval keyword is proper, an OK button is selected (Y at step S9) and then a retrieving engine performs full-text retrieval based upon the retrieval keyword from text data files in a database (step S10) and displays retrieval results on a console panel (step S11). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an information processing apparatus that performs file search and the like, an image forming apparatus, a program, and a storage medium that stores the program.

  Patent Document 1 discloses a technique for removing noises such as acoustic echoes from voices input to a microphone when human voices, background noises, and the like are mixed in the acoustic echoes generated by voices output from speakers. Yes.

JP 10-257583 A

  Digital copying machines are known that have a function of storing files such as image data and a network function. In such a digital copying machine, the stored file also stores text data obtained by OCR processing of a PDF file read by a scanner, text data received from the outside via the Internet or LAN, and the like. Therefore, if a full-text search can be performed on a file containing such accumulated text data, it is convenient to find a necessary file in advance. Various techniques for such full-text search are well known.

  However, if a search keyword is input by keyboard operation when performing this full-text search, the full-text search function is not easy to use, and the full-text search can be performed by a simpler operation.

  An object of the present invention is to facilitate the input of a search keyword and improve the usability when searching for a file group.

  The present invention includes a microphone, speech recognition means for recognizing a voice input through the microphone and converting it into text data, and search means for searching a predetermined file group using the text data as a search keyword. Information processing apparatus.

  Another aspect of the present invention is a scanner that reads an image of a document, a printer that forms an image on a medium based on the image data after the reading, a storage unit that stores a file group, a microphone, The image forming apparatus includes speech recognition means for recognizing voice converted by the microphone and converting it into text data, and search means for searching the file group using the text data as a search keyword.

  Another aspect of the present invention relates to a voice recognition means for recognizing a voice inputted by a predetermined microphone and converting it into text data, and a search means for searching a predetermined file group using the text data as a search keyword. Is a computer-readable program that causes a computer to execute.

  According to the present invention, since the search keyword can be input by voice input without using a keyboard or the like, the usability of the search function can be improved.

  The best mode for carrying out the present invention will be described.

  In this embodiment, as an information processing apparatus and an image forming apparatus, a copy function, a facsimile (FAX) function, a print function, a scanner function, and an input image (an original image read by the scanner function or an image input by a printer or a FAX function) An example applied to a so-called digital color multi-function peripheral that combines functions for distributing the image and the like.

  FIG. 1 is a system configuration diagram including a digital color multifunction peripheral 1 according to the present embodiment. As shown in FIG. 1, in the present embodiment, a server computer 3 or a plurality of generations of client computers that execute various kinds of information processing via a local area network (LAN) 2 as a communication network is connected to the digital color MFP 1. Assume a system in which 4 is connected. The server computer 3 supports, for example, the FTP and HTTP protocols, and realizes functions of a Web server and DNS (Domain Name Server). That is, in this system, there is an environment in which image processing functions such as an image input function (scanner function), an image output function (print function), and an image storage function provided in the digital color multifunction peripheral 1 can be shared on the LAN 2. It has been built.

  Such a system is connected to the Internet network 6 via the communication control unit 5 and is constructed so as to be able to communicate data with an external environment via the Internet network 6. The communication control unit 5 is generally a router, an exchange, a modem, a DSL modem, or the like, but it is sufficient that TCP / IP communication is possible at a minimum. The LAN 2 is not limited to wired communication, but may be wireless communication (infrared rays, radio waves, etc.). Further, an optical fiber may be used.

  Next, the digital color multifunction peripheral 1 will be described. Here, FIG. 2 is an external perspective view schematically showing the digital color multifunction peripheral 1, and FIG. 3 is a block diagram showing electrical connection of each part of the digital color multifunction peripheral 1. As shown in FIG. As shown in FIG. 2, the digital color multifunction peripheral 1 includes an image reading device 8 that reads an image from a document on an upper portion of a printing device 7 (printer) that is an image forming device that forms an image on a medium such as transfer paper. It is set as the set-up. An operation panel P that allows various inputs such as display to the operator and function setting from the operator is provided on the outer surface of the image reading device 8 (scanner). Further, under the operation panel P, program codes and image data stored in a storage medium M (see FIG. 3) such as an optical disk and a flexible disk are read, or program codes and images are read from the storage medium M. An external media input / output device 9 that is a device for writing data and the like is provided with an insertion opening that allows insertion of the storage medium M exposed to the outside.

  As shown in FIG. 3, the structure of the digital color multifunction peripheral 1 is roughly divided into an image processing unit A and an information processing unit B, and the printing device 7 and the image reading device 8 are image processing units. The operation panel P and the external media input / output device 9 belong to the information processing unit B.

  The image processing unit A will be described. The image processing unit A including the printing device 7 and the image reading device 8 includes an image processing control unit 10 that performs overall control of image processing in the image processing unit A. The image processing control unit 10 includes a printing unit. A print control unit 11 that controls the apparatus 7 and an image reading control unit 12 that controls the image reading apparatus 8 are connected.

  The print control unit 11 outputs a print instruction including image data to the printing apparatus 7 under the control of the image processing control unit 10, and causes the printing apparatus 7 to form and output an image on a medium such as transfer paper. The printing device 7 is capable of full-color printing. In addition to the electrophotographic method, the printing method includes various methods such as an inkjet method, a sublimation type thermal transfer method, a silver salt photography method, a direct thermal recording method, and a melt type thermal transfer method. Can be used.

  The image reading control unit 12 drives the image reading device 8 under the control of the image processing control unit 10, and reflects the reflected light of the lamp irradiation on the surface of the document to a light receiving element (for example, a CCD (Charge Coupled Device)) using a mirror and a lens. Light is collected and read, and A / D converted to generate 8-bit RGB digital image data.

  Such an image processing control unit 10 includes a central processing unit (CPU) 13 as a main processor and an SDRAM (Synchronous) that temporarily stores image data read from the image reading device 8 for image formation by the printing device 7. Dynamic Random Access Memory) 14, ROM (Read Only Memory) 15 storing a control program, NVRAM 16 capable of holding data even when the power is turned off to record system log / system setting / log information, The microcomputer is configured with a bus connection.

  Further, the image processing control unit 10 includes an image processing unit via an HDD (Hard Disk Drive) 17 serving as a storage device for storing a large amount of image data and a job history, and a HUB 19 which is a concentrator provided inside the device. A LAN control unit 18 for connecting the part A to the LAN 2 and a FAX control unit 20 for performing FAX control are connected. The FAX control unit 20 is connected to a private branch exchange (PBX) 22 that communicates with the public telephone network 21, and the digital color multifunction peripheral 1 can communicate with a remote facsimile apparatus.

  In addition, a display control unit 23 and an operation input control unit 24 are connected to the image processing control unit 10. The display control unit 23 outputs an image display control signal to the information processing unit unit B via the communication cable 26 connected to the control panel I / F 25 under the control of the image processing control unit 10. The image display is controlled on the operation panel P. Further, the operation input control unit 24 connects an input control signal corresponding to the function setting or input operation by the operator from the operation panel P of the information processing unit B to the control panel I / F 25 by the control of the image processing control unit 10. Input via the communication cable 26. In other words, the image processing unit A is configured to be able to directly monitor the operation panel P of the information processing unit B via the communication cable 26.

  Accordingly, the image processing unit A is configured such that the communication cable 26 is connected to the image processing unit provided in the conventional image processing apparatus, and the operation panel P of the information processing unit B is used. That is, the display control unit 23 and the operation input control unit 24 of the image processing unit A operate as if they are connected to the operation panel P.

  With such a configuration, the image processing unit A analyzes the print data that is image information from the outside (the server computer 3, the client computer 4, the facsimile machine, etc.) and the command that instructs printing, and outputs the print data to the output image data The bitmap is expanded so that it can be printed, and the printing mode is analyzed from the command to determine the operation. The print data and command are received and operated through the LAN control unit 18 or the FAX control unit 20.

  In addition, the image processing unit A externally print data stored in the SDRAM 14 or the HDD 17, document reading data, output image data processed for output, and compressed data obtained by compressing them externally (server computer 3, Client computer 4, facsimile, etc.).

  Further, the image processing unit A transfers the read image data of the image reading device 8 to the image processing control unit 10, corrects signal deterioration due to quantization to the optical system and digital signals, and stores the image data in the SDRAM 14. Write. The image data stored in the SDRAM 14 in this way is converted into output image data by the print control unit 11 and output to the printing apparatus 7.

  Next, the information processing unit B including the operation panel P will be described. The information processing unit part B has a microcomputer configuration controlled by a general-purpose OS (Operating System) used in an information processing apparatus generally called a personal computer. The information processing unit section B has a CPU 31 that is a main processor, and the CPU 31 includes a memory unit that includes a ROM that is a read-only memory that stores a RAM that is a work area of the CPU 31, a startup program, and the like. 32 and a storage device control unit 35 that controls input / output of data to / from a storage device 34 such as an HDD that stores an OS (Operating System) and application programs are connected via a bus.

  The CPU 31 is connected to a LAN control unit 33 that is a communication interface for connecting the information processing unit unit B to the LAN 2 via the HUB 19. The IP address that is the network address assigned to the LAN control unit 33 is different from the IP address assigned to the LAN control unit 18 of the image processing unit A described above. That is, two IP addresses are assigned to the digital color multifunction peripheral 1 of the present embodiment. That is, the image processing unit unit A and the information processing unit unit B are each connected to the LAN 2, and the image processing unit unit A and the information processing unit unit B can exchange data. ing.

  Since the digital color multifunction peripheral 1 is connected to the LAN 2 via the HUB 19, it appears that only one IP address is assigned. Therefore, the aesthetic appearance is not impaired, and handling such as connection can be facilitated.

  Further, a display control unit 36 and an operation input control unit 37 that control the operation panel P are connected to the CPU 31. Here, FIG. 4 is a plan view showing the configuration of the operation panel P. FIG. As shown in FIG. 4, the operation panel P includes a display device 40 that is, for example, an LCD (Liquid Crystal Display) and an operation input device 41. The operation input device 41 includes an ultrasonic acoustic wave type touch panel 41a laminated on the surface of the display device 40 and a keyboard 41b having a plurality of keys. The keyboard 41b is provided with a start key for declaring the start of image reading, a numeric keypad for inputting numerical values, a reading condition setting key for setting the transmission destination of the read image data, a clear key, and the like. That is, the display control unit 36 outputs an image display control signal to the display device 40 via the control panel I / F 38 and causes the display device 40 to display predetermined items corresponding to the image display control signal. On the other hand, the operation input control unit 37 receives, via the control panel I / F 38, an input control signal corresponding to the function setting or input operation by the operator in the operation input device 41.

  In addition, a control panel communication unit 39 connected to the control panel I / F 25 of the image processing unit A via the communication cable 26 is connected to the CPU 31. The control panel communication unit 39 receives the image display control signal output from the image processing unit unit A, and receives the input control signal according to the function setting or input operation by the operator from the operation panel P. Forward to. Although details will be described later, the image display control signal from the image processing unit A received by the control panel communication unit 39 is subjected to data conversion processing for the display device 40 of the operation panel P and then output to the display control unit 36. The input control signal corresponding to the function setting or input operation by the operator from the operation panel P is input to the control panel communication unit 39 after being subjected to data conversion processing in a format according to the specifications in the image processing unit A. .

  As described above, the storage device 34 stores an OS (Operating System) and application programs executed by the CPU 31. In this sense, the storage device 34 functions as a storage medium that stores application programs. In the digital color multifunction peripheral 1, when the user turns on the power, the CPU 31 activates the activation program in the memory unit 32, reads the OS from the storage device 34 into the RAM in the memory unit 32, and activates the OS. Such an OS activates a program, reads information, and stores information in accordance with a user operation. As a representative OS, Windows (registered trademark) and the like are known. These operation programs running on the OS are called application programs. The OS of the information processing unit section B is the same type of OS as the information processing apparatus (server computer 3, client computer 4, etc.), that is, a general-purpose OS (for example, Windows (registered trademark)).

  As described above, the digital color multi-function peripheral 1 of the present embodiment has a storage medium M that stores various program codes (control programs) such as an OS, device drivers and various application programs, image data, and the like. , Flexible disk, hard disk, optical disk (CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-R, DVD + R, DVD-RW, DVD + RW, etc.), magneto-optical disk (MO), semiconductor A flexible disk drive device, an optical disk drive device, and an MO drive device, which are devices that read program codes, image data, and the like stored in a storage medium M such as a medium, or write program codes, image data, and the like to the storage medium M , Media drive devices, etc. External media input-output device 9 is mounted. Such an external media input / output device 9 is controlled by an input / output device control unit 42 connected to the CPU 31 by a bus.

  Therefore, the application program stored in the storage device 34 may be one in which the application program recorded in the storage medium M is installed. Therefore, the storage medium M can also be a storage medium that stores application programs. Furthermore, the application program may be imported from the outside via, for example, the Internet network 6 and the LAN 2 and installed in the storage device 34.

  The input / output device control unit 42 is also connected with various interfaces 43 such as USB, IEEE 1394, SCSI, and the like, and various devices (such as digital cameras) can be connected via the various interfaces 43.

  Next, the contents of characteristic processing executed by the digital color multifunction peripheral 1 will be described. This processing is executed by a search execution program that is read from the storage medium M by the external media input / output device 9 and installed in the storage device 34.

  FIG. 5 is a block diagram showing a system configuration of a system executed by the search execution program.

  The information processing unit B of the digital color multifunction peripheral 1 includes a microphone 101 that accepts voice input. For example, the microphone 101 is provided on the operation panel P (see FIG. 4). The microphone 101 can take in audio data to the digital color multifunction peripheral 1 via a predetermined interface 102 for audio input serving as various interfaces 43. The input voice data is stored in the voice storage unit 103. Specifically, the voice storage unit 103 is a predetermined storage area of the memory unit 32 or the like.

  The database 104 (storage means) is constructed in the storage device 34 and the like, and various files such as text data files are registered. Specifically, this file is processed by OCR (Optical Character Reader) processing on data received from the outside such as the Internet 6, the server computer 3 and the client computer 4 via the LAN 2 or image data read by the image reading device 8. Is converted to text (for example, data converted to text by performing OCR processing on a PDF file).

  The voice recognition unit 105 converts voice data stored in the voice storage unit 103 into text data by a known technique. Therefore, the voice spoken by the user toward the microphone 101 becomes text data in the voice recognition unit 105.

  The search engine 106 performs a full-text search for a text data file group registered in the database 104 using a given search keyword. Various systems are known as specific means for the full-text search, and detailed description thereof is omitted. In this case, the search engine 106 can use the text data converted by the voice recognition unit 105 as a search keyword. In the above system configuration, the voice recognition unit 105 and the search engine 106 are realized by processing executed by the CPU 31 based on the above-described search execution program.

  Next, processing executed using the system of FIG. 5 will be described with reference to the flowchart of FIG.

  The user selects the search button 111 from the menu screen 141 displayed on the operation panel P shown in FIG. 7 when searching the entire text data file group in the database 104 (Y in step S1). Thereby, the search screen 142 shown in FIG. 8 is displayed (step S2). When the keyboard input button 112 is selected on this search screen, a keyboard is displayed on the touch panel of the operation panel P, and a search keyword for full-text search can be input with the keyboard. Here, the voice input button 113 is selected. While the voice input button 113 is being pressed (Y in Step S3), voice is captured from the microphone 101 and voice data is stored in the voice storage unit 103 (Step S4). Therefore, the user can input the search keyword without operating the keyboard by speaking to the microphone 101 the word to be used as the search keyword while pressing the voice input button 113.

  When the search keyword is input by voice and the voice input button 113 is not pressed (Y in step S5), the input search keyword is voice-recognized by the voice recognition unit 105 and converted into text data ( (Speech recognition means) (step S6) and displayed in the keyword display field 144 of the search keyword display screen 143 shown in FIG. 9 (step S7). When the displayed search keyword is not appropriate, if the re-input button 114 is selected (Y in step S8), the process returns to step S2. If the displayed search keyword is considered appropriate, if the OK button 115 is selected (Y in step S9), the search engine 106 performs a full text search of the text data file in the database 104 based on the search keyword ( Search means) (step S10), and the search result is displayed on the operation panel P (step S11).

  As described above, since the search keyword can be input by voice, a full-text search of the database 104 is easier than in the case of keyboard operation.

  By the way, since such a full-text search system is installed in the digital color multi-function peripheral 1, noise generated from the mechanism section of the digital color multi-function peripheral 1 becomes an obstacle, and the recognition rate of voice recognition by the voice recognition section 105 is lowered. I will let you. Therefore, the speech recognition unit 105 takes measures against noise. Below, the process performed as this countermeasure is demonstrated.

  That is, a remote speaker is usually used in a video conference system, a hands-free car phone system, and a car navigation system, and an audio / acoustic signal output from the speaker is a room wall or desk in the video conference system. It is reflected on the window glass and dashboard of the car and mixed into the microphone input as an acoustic echo. Since this acoustic echo becomes a noise component with respect to the input of the microphone, it is desirable to eliminate it from the microphone input.

  As countermeasures for this, the LMS (Least Means Square) method, the NLMS (Normalized Least Mean Square) method, and the like have been proposed. These technologies are described in “Chapter 4 of Corona,” edited by Nobuhiko Kitawaki, “Sound Communication Engineering-Speech and Acoustic Technology in the Multimedia Age”.

  In any of the LMS method, the NLMS method, and their improved methods, a pseudo acoustic echo generated from the estimated value of the impulse response of the transmission path of the acoustic echo and the source signal of the sound / acoustic signal output from the speaker Based on the principle of adaptive filtering, the basic principle is to dynamically and sequentially adapt the impulse response of the acoustic echo transmission path so that the difference between the input signal and the microphone input signal becomes smaller.

  Therefore, if a signal other than an acoustic echo, such as a human voice or background noise, is mixed in the microphone input, the impulse response of the acoustic echo transmission path is an ideal value due to these effects. The phenomenon of deviating from occurs. This state is called simultaneous conversation (double talk). In the double talk state, the adaptive operation of the impulse response of the acoustic echo transmission path is stopped, and the impulse response at this time is held. When no sound is generated, the impulse response held immediately before the stop is restored to prevent the acoustic echo cancellation performance from significantly decreasing.

  Therefore, the speech recognition unit 105 includes the speech processing system 121 shown in FIG. 10 that performs such noise cancellation. Hereinafter, the voice processing system 121 will be described. In FIG. 10, reference numeral 131 denotes an adaptive filter (adaptive filter means) that generates a pseudo acoustic echo signal based on the acoustic signal and the updated impulse response. The subtraction result of the subtraction unit 133 (subtraction means), that is, the echo A removal signal (a signal after removing the noise component) is input, and the filter coefficient is sequentially updated so that the level of the echo removal signal becomes 0 (zero). The adaptive filter 131 receives an on / off stop / restart signal from a VAD (Voice Activity Detection) 132 (VAD means), which will be described later, and adapts the impulse response when the stop is instructed, that is, when the above-described double talk occurs. The filter coefficient at the time when the operation is stopped and the stop signal instructed to recover is input is held.

  The subtracting unit 133 subtracts the pseudo echo signal from the input signal input from the microphone 101 and eliminates a noise component from the input signal.

  VAD132 is a system that discriminates whether the input signal is speech (human voice) or non-speech. This system is written by Sadaaki Furui, “Digital Speech Processing” (Page 153 of Tokai University Press) and “Recommendation GSM 06.32” The adaptive filter 131, the subtracting unit 133, and the VAD 132 can be realized by wired logic, but here, description will be made assuming that they are realized by software based on the above-described search execution program.

  When the VAD 132 determines that an audio signal is included in the input signal, an impulse response adaptive operation stop signal is output from the VAD 132 to the adaptive filter 131, and when the VAD 132 determines that the input signal is non-speech, the impulse response An adaptive operation restart signal is output to the adaptive filter 131.

  The noise component of the sound input to the microphone 101 is the sound component of the noise emitted from the digital color copying machine 1 reflected by the wall of the room where the digital color copying machine 1 is installed and input to the microphone 101 (acoustic echo, 11 (b)) and background noise (see FIG. 11 (c)) specific to the usage environment. The acoustic signal input to the adaptive filter 131 is the operation sound when the digital color copying machine 1 emits an operation sound when a predetermined operation is performed, or the sound when the sound is emitted by voice guidance. (Output from the speaker 107 (see FIG. 4)).

  In the system configuration as shown in FIG. 10, in a state where no sound is input from the microphone 101, an acoustic echo and background noise reproduced and output from the speaker 107 are input from the microphone 101. When the echo cancellation signal obtained by subtracting the pseudo acoustic echo from the input signal by the subtracting unit 133 is non-speech, the VAD 132 outputs a restart signal while the VAD 132 is determining. While updating the filter coefficient so that the echo cancellation signal becomes 0 (zero), a pseudo echo signal for canceling the acoustic echo component of the input signal is generated. As shown in FIG. 11, when a voice is input at timing t1 and the VAD 132 determines that a voice is included in the input signal at timing t2, a stop signal is sent to the adaptive filter 131. As a result, the adaptive filter 131 stops updating the filter coefficient, and fixes its own filter coefficient value to the filter coefficient value at the timing t2. As a result, only the acoustic echo component in the input signal composed of the speech component and the noise component (acoustic echo and background noise) input from the microphone 101 is erased, and the speech recognition unit 105 performs the above-described speech recognition unit 105 for the speech signal after the erasure. Voice recognition is performed as follows.

  As described above, when inputting a search keyword by voice, noises such as operation sounds and background noise are removed from the input voice, and voice recognition is performed on the voice after the removal. Can improve the voice recognition rate.

1 is a system configuration diagram including a digital color multifunction peripheral according to an embodiment of the present invention. 1 is an external perspective view schematically showing a digital color multifunction peripheral. FIG. 3 is a block diagram showing electrical connections of each part of the digital color multifunction peripheral. It is a top view which shows the structure of an operation panel. It is a block diagram which shows the system configuration | structure of the system performed by a search execution program. It is a flowchart of the process performed with the system of FIG. It is a top view of a menu screen. It is a top view of a search screen. It is a top view of a search keyword screen. It is a block diagram of the voice processing system with which a voice recognition part is provided. It is explanatory drawing of the process which a speech processing system performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Information processing apparatus, image forming apparatus 7 Printer 8 Scanner 121 Storage means 131 Adaptive filter means 132 VAD means 133 Subtraction means

Claims (7)

With a microphone,
Speech recognition means for recognizing voice converted by the microphone and converting it into text data;
Search means for searching a predetermined file group using the text data as a search keyword;
An information processing apparatus comprising: Speakers,
Adaptive filter means for generating a pseudo acoustic echo signal from the estimated value of the impulse response of the acoustic echo transmission path of the sound emitted by the speaker and the acoustic signal output from the speaker;
Subtracting means for subtracting the acoustic echo signal from the speech signal input by the microphone and before the speech recognition to obtain an echo removal signal;
When it is determined that an audio signal is included in the echo cancellation signal, the adaptive filter unit is instructed to stop the impulse response adaptive operation. When it is determined that no audio signal is included, the impulse response adaptive operation is performed. VAD (Voice Activity Detection) means for instructing resumption,
The information processing apparatus according to claim 1, further comprising: A scanner that reads the image of the document,
A printer that forms an image on a medium based on the read image data;
Storage means for storing files,
With a microphone,
Speech recognition means for recognizing voice converted by the microphone and converting it into text data;
Search means for searching the file group using the text data as a search keyword;
An image forming apparatus. Speakers,
Adaptive filter means for generating a pseudo acoustic echo signal from the estimated value of the impulse response of the acoustic echo transmission path of the sound emitted by the speaker and the acoustic signal output from the speaker;
Subtracting means for subtracting the acoustic echo signal from the speech signal input by the microphone and before the speech recognition to obtain an echo removal signal;
When it is determined that an audio signal is included in the echo cancellation signal, the adaptive filter unit is instructed to stop the impulse response adaptive operation. When it is determined that no audio signal is included, the impulse response adaptive operation is performed. VAD (Voice Activity Detection) means for instructing resumption,
The image forming apparatus according to claim 3, further comprising: Voice recognition means for recognizing a voice input with a predetermined microphone and converting it into text data;
Search means for searching a predetermined file group using the text data as a search keyword;
A computer-readable program that causes a computer to execute. Adaptive filter means for generating a pseudo acoustic echo signal from the estimated value of the impulse response of the acoustic echo transmission path of the sound emitted by the predetermined speaker and the acoustic signal output from the speaker;
Subtracting means for subtracting the acoustic echo signal from the speech signal input by the microphone and before the speech recognition to obtain an echo removal signal;
When it is determined that an audio signal is included in the echo cancellation signal, the adaptive filter unit is instructed to stop the impulse response adaptive operation. When it is determined that no audio signal is included, the impulse response adaptive operation is performed. VAD (Voice Activity Detection) means for instructing resumption,
The program according to claim 5, further causing the computer to execute. A storage medium storing the program according to claim 5 or 6.

Images