JP2020089641A - Voice recognition input device, voice recognition input program, and medical image capturing system - Google Patents

Abstract

To allow an operation instruction by an operator's voice to be recognized accurately.SOLUTION: Provided is a voice recognition input device in which a voice recognition data table is stored in which a plurality of voice commands are associated with one operation command and a weighting coefficient corresponding to the frequency of use of the voice command is recorded for each voice command, voice recognition processing is performed with voice input as a recognition target, the voice command is output as a result of the voice recognition processing, and the voice command is converted into the operation command recorded corresponding to the voice command referring to the voice recognition data table to output the operation command to an external device. In this case, a plurality of voice command candidates that can correspond to voice input are selected and each of the plurality of voice command candidates is multiplied by the weighting coefficient, thereby outputting the most probable voice command as a result of the voice recognition processing with respect to the voice input.SELECTED DRAWING: Figure 1

Description

The present invention relates to a voice recognition input device, a voice recognition input program, and a medical image capturing system, and more particularly, to a voice recognition input device that is connected to a medical device such as a medical image capturing device and outputs a command to the medical device. The present invention relates to a recognition input program and a medical image capturing system.

In recent years, with the advancement of interventional treatment, there are an increasing number of cases in which blood vessels and digestive tracts are inspected or treated under fluoroscopic imaging and under flexible endoscope operation. It is necessary to ensure the cleanliness of the operator in these examinations or treatments, and manual operation of the equipment cannot maintain the cleanliness.
In addition, for example, in an examination using an X-ray fluoroscopic imaging apparatus, a medical staff supporting the operator often operates the device under the oral instruction of the operator who performs the examination procedure. In such a case, it may take some time for the operator to communicate with the medical staff as an assistant, such as oral instructions, and it may take time to operate the device as intended by the operator. In addition, it is desirable to operate the medical device by direct oral instruction.
Therefore, in medical devices used for surgery and examinations, operation by voice recognition is desired to ensure cleanliness and improve operability.

On the other hand, in recent years, in speech recognition technology, in addition to the existing method using a hidden Markov model, a method using Deep Learning appears, which enables not only word recognition but also speech recognition processing as a sentence. Is improving. In addition, some speech recognition processes perform large-scale machine learning using a server or Cloud to sequentially improve performance, but medical devices need to be designed in consideration of confidentiality. A voice recognition input device applied to a medical device needs to perform voice recognition processing in a non-network environment without connecting to a Cloud or a server.

As an example of a medical device that operates by voice recognition, in Patent Document 1, in an X-ray image diagnostic apparatus, in order to reduce the burden on the operator, there is a risk that the test subject may be harmed by a malfunction. It is disclosed that a function is controlled based on a manual operation by an operator, and a function that does not cause harm to a subject even by a malfunction is controlled by recognizing a voice generated by the operator. ..

JP 2006-149909 A

However, if a plurality of voice operation commands having phonemes similar to each other are registered in the database used for the voice recognition process, there is a possibility that erroneous detection may occur in the voice recognition process. That is, it is not possible to determine which voice operation command among the plurality of voice operation commands composed of similar phonemes the voice uttered by the operator corresponds to, which may result in an erroneous detection. In this case, the operation cannot be performed by the voice recognition processing, and as a result, the surgeon instructs the medical staff who provides support to operate the equipment, which shortens the time required for the equipment to operate as intended by the operator. I can't. Further, in that case, it is necessary to improve the recognition rate of the voice operation command, but the means is not specified.

The present invention has been made in view of the above circumstances, and an object of the present invention is to accurately recognize an operation instruction by an operator's voice in voice recognition processing and reduce false detection.

In order to solve the above problems, the present invention provides the following means.
One aspect of the present invention is a voice recognition input device for inputting an operation command to an external device, wherein a plurality of voice commands are recorded in association with one operation command, and the voice command is used for each voice command. A storage unit storing a voice recognition data table in which weighting factors corresponding to frequencies are stored, a voice input is accepted, a voice recognition process is performed with the voice input as a recognition target, and a voice command corresponding to the voice input is voice recognized. A voice recognition unit that outputs as a result, a command conversion unit that refers to the voice recognition data table and converts the voice command into an operation command recorded corresponding to the voice command, and the operation command by the external unit. An operation determining unit for outputting to a device, the voice recognition unit selects a plurality of voice command candidates that can correspond to the voice input, and multiplies each of the plurality of voice command candidates by the weighting factor, which is the most reliable. There is provided a voice recognition input device for outputting a new voice command as a result of voice recognition processing for the voice input.
According to the present invention, since the voice recognition processing is performed using the data table in which the weighting coefficient corresponding to the usage frequency is recorded for each voice command, it is possible to improve the accuracy of the voice recognition processing in a voice operation instruction.

According to the present invention, it is possible to reduce erroneous detection in the voice recognition processing and accurately recognize the operation instruction by the operator's voice.

It is a block diagram showing a schematic structure of a voice recognition input device concerning a 1st embodiment of the present invention. 3 is an example of a voice recognition data table stored in a voice recognition DB of the voice recognition input device in FIG. 1. 6 is a graph showing the relationship among the operation command use frequency, the operation command code, and the offset coefficient T, which is referred to when updating the weight coefficient in the voice recognition input device according to the first exemplary embodiment of the present invention. It is a flowchart which shows the flow of the voice recognition input process by the voice recognition input device which concerns on the 1st Embodiment of this invention. 7 is a graph showing a relationship among the number of times a voice command is integrated, a voice command code, and an offset coefficient V, which is referred to when updating a weighting factor, in the voice recognition input device according to the modified example of the first exemplary embodiment of the present invention. An example of the data table stored in speech recognition DB in the speech recognition input device which concerns on the 2nd Embodiment of this invention is shown, (A) is a data table which shows the state of inspection start information, (B) is inspection. A data table showing the status of the type, (C) is a data table showing the status of the X-ray irradiation information, and (D) is a data table showing the apparatus operation status and the classification of the voice recognition DB. 8 is a flowchart showing a flow of a voice recognition data table switching process in the voice recognition input device according to the second embodiment of the present invention.

The voice recognition input device according to the embodiment of the present invention inputs data to an external device such as a medical image capturing device connected to the voice recognition input device.

(First embodiment)
Hereinafter, the voice recognition input device according to the first embodiment of the present invention will be described in more detail with reference to the drawings. FIG. 1 shows a schematic configuration diagram of a voice recognition input device according to the present embodiment.
The voice recognition input device 10 includes a central processing unit (CPU) 11 that controls the voice recognition input device 10 as a whole, a voice input I/F (interface) 12 that receives voice input such as a microphone, a mouse, a keyboard, and the like for manual input. A manual input I/F (interface) 13, a memory 14, a voice recognition DB 15 that stores voice recognition algorithms and data necessary for voice recognition processing, and a log collection DB 16 that collects and records logs related to voice input are provided. The components are connected to each other via the system bus.

In the present embodiment, the voice recognition input device 10 is communicably connected to the image pickup device 20 and gives various input instructions to the image pickup device 20. Further, the voice recognition input device 10 is connected to the display 30 via the image capturing device 20, displays the image or the like acquired by the image capturing device 20 on the display 30, and enlarges or reduces the displayed image. Give desired operation instructions. As the image capturing device 20, hardware for acquiring a medical image such as an X-ray device, an MRI device, a CT device, and a PET device can be applied.

As shown in FIG. 1, the CPU 11 includes a voice operation processing unit 120, a manual operation processing unit 130, and a system operation determining unit 140 in order to input a voice input instruction to the image capturing apparatus or the like by the voice recognition input device 10. Realize the function. In particular, the voice operation processing unit 120 realizes the functions of the voice recognition unit 111, the command conversion unit 112, and the command analysis unit 113.

The functions of these units realized by the CPU 11 can be realized as software by the CPU previously reading and executing a program stored in a memory such as a magnetic disk (not shown). Note that part or all of the operations executed by the respective units included in the CPU 11 can be realized by an ASIC (application specific integrated circuit) or an FPGA (field-programmable gate array).

The voice operation processing unit 120 recognizes a voice operation instruction (voice input) input via a voice input I/F 12 such as a microphone and gives an operation instruction to the image capturing apparatus 20. The functions of the recognition unit 111, the command conversion unit 112, and the command analysis unit 113 are realized.

The voice recognition unit 111 performs voice recognition processing on a voice operation instruction input via the voice input I/F 12 according to a voice recognition algorithm stored in advance in the voice recognition DB 15 or the like, and outputs a voice command as a recognition result. Is output to the command conversion unit 112. Here, in the voice recognition process, the voice recognition unit 111 performs the voice recognition process according to a predetermined voice recognition algorithm using a voice recognition data table (see FIG. 2) stored in a voice recognition DB 15 described later, and outputs a voice as a recognition result. Select a command. Details of the voice recognition processing by the voice recognition unit 111 will be described later.

The command conversion unit 112 converts an operation command corresponding to the voice command selected through the voice recognition processing in the voice recognition unit 111, and the operation command is used as an operation instruction by the operator, the system operation determination unit 140 and the command analysis unit 113. Output to.

The command analysis unit 113 acquires information on the operation command related to the operation instruction from the operator from the command conversion unit 112, generates an operation history and records it in the log collection DB 16, and analyzes the operation history. Here, the information regarding the operation command may include not only the operation command but also a voice command before being converted into the operation command. Further, the command analysis unit 113 updates the weighting coefficient of the voice recognition data table based on the analysis result of the operation history. Details of updating the weighting factor will be described later.

When performing a manual operation, the manual operation processing unit 130 generates an operation command related to the operation instruction based on the operation instruction input via the manual input I/F 13 such as a mouse or a keyboard, and performs a system operation. Output to the determination unit 140.
The system operation determination unit 140 outputs the operation command input from the voice operation processing unit 120 or the manual operation processing unit 130 to the image capturing device 20 and the command analysis unit 113.

The voice input I/F 12 receives a speech of an operator such as an operator as a voice operation instruction (voice input), converts it into voice data of an electric signal, and outputs the voice data to the voice operation processing unit 120. For example, a microphone or the like can be applied.
The manual input I/F 13 receives a manual operation instruction from an operator, converts the received operation instruction into an electric signal, and outputs the electric signal to the manual operation processing unit 130. For example, a mouse, a keyboard, a touch panel, or the like. An input device can be applied.

The memory 14 stores the program executed by the CPU 11 and the progress of arithmetic processing, and temporarily stores voice and manual operation instructions.
The voice recognition DB 15 stores a predetermined voice recognition algorithm and also records a voice recognition data table used for voice recognition processing. Details of the voice recognition data table will be described later.
The log collection DB 16 acquires and records information related to operation instructions such as an operation history generated by the command analysis unit 113.

(About voice recognition processing and voice recognition data table)
As shown in FIG. 2, the voice recognition data table includes voice command data db1 indicating data regarding voice commands, operation command data db2 indicating data regarding operation commands, and weighting factors indicating weighting factors determined for each voice command. It is composed of data db3.

As shown in FIG. 2, in the voice recognition data table, a plurality of voice commands db11 are recorded in association with the same operation command db21. By doing so, even if the operator gives different habits or utterances to the same operation instruction, the same operation can be realized through the voice recognition process. Also, a command reading db12 and a voice command code db13 are recorded in association with each voice command of the voice command data db1. Regarding the operation command db21, similarly, an operation command code db22 is associated with each operation command and stored.

By the way, the voice recognition unit 111 performs a voice recognition process as follows. That is, the voice recognition unit 111 first converts the voice data input via the voice input I/F into sound waves, and decomposes and identifies each character of the voice data from the sound waves into phonemes. Then, the phoneme of the voice data is compared with the phoneme of the command reading db12 using a voice recognition algorithm such as statistical machine learning according to the hidden Markov model or machine learning using a deep learning model. By this matching, the voice recognition unit 111 selects a command reading db12 similar to the voice data, and outputs a candidate of the voice command db11 corresponding to the selected command reading and a score as an index of its certainty.

Here, as an example of the voice recognition processing, consider a case where the operator speaks “image reduction” and gives an operation instruction. When the operator speaks “image reduction”, the phonemes of “gazoshushu” and “gazoshushu” halfway match in the command reading db12 of the voice recognition data table. Therefore, the voice recognition unit 11 may erroneously recognize "gazoshusho" as "image collection" depending on the quality of the input voice data. In this case, the operator selects the "perspective recording" of the operation command db21 corresponding to the "gazoshushu" of the voice command db11 as the operation command due to the incorrect recognition of the voice recognition, although the operator has instructed the image reduction of the operation command db21. Then, an operation unintended by the operator is performed.

Therefore, in the voice recognition data table, the weighting factor data db3 is recorded in association with each voice command in order to avoid such erroneous recognition. The weighting factor data db3 is a weighting factor for multiplying a score attached to each of one or more voice command candidates output in the process of voice recognition processing. The weighting factor recorded in the weighting factor data db3 is a value corresponding to each voice command and determined according to the frequency of use of the voice command and the like.

Further, the weighting factor can be updated based on the result of the command analysis unit 113 analyzing the operation history recorded in the log collection DB 16. That is, the command analysis unit 113 analyzes the operation history to calculate the number of times of use for each operation command within a certain period, and calculates the offset coefficient T for updating the weighting coefficient based on the number of times of use.

The offset coefficient T can be determined by, for example, the cumulative number of times the operation command code corresponding to each operation command is issued according to the graph shown in FIG. In addition, the offset coefficient T can be calculated based on the ratio of each operation command to the total number of operation command codes issued in a predetermined period. The command analysis unit 113 updates the weight coefficient by multiplying each calculated weight coefficient by the calculated offset coefficient T corresponding to the operation command.

The analysis of the operation history by the command analysis unit 113 can be automatically performed, or can be performed at an arbitrary timing such as according to an operation instruction from an operator or a device provider. In addition, the command analysis unit 113 can update the weighting factor by subtracting from the number of times the operation command has been used, if an erroneous detection occurs in the voice recognition process of the voice recognition unit 111.

The flow of voice input processing by the voice recognition input device 10 configured as described above will be described below with reference to the flowchart of FIG.
As shown in FIG. 4, when the voice recognition input device 10 operates, it enters a standby state for voice input. When there is a voice input in the voice input I/F (step S101), the process proceeds to step S102, where the voice recognition unit 111 accepts the voice data input from the voice input I/F 12 and the phoneme of the voice data and the voice recognition DB 15. The phoneme of each command reading registered in is compared and the voice command candidate recorded corresponding to the command reading is selected. The selection of this candidate can be judged based on a score that is an index of the certainty of voice recognition.

In the next step S103, it is determined whether or not there are one or more voice command candidates selected in step S102. If the number of voice command candidates is not one or more, the process proceeds to step S104 and it is determined that there is no voice command in step S101. Return to. If the number of voice command candidates is 1 or more, the process proceeds to step S105, and the score assigned to each voice command candidate and the weighting coefficient recorded in the voice recognition data table corresponding to the voice command are set. To multiply.

In step S106, the voice command candidate with the highest score is selected as a result of the multiplication of the score and the weighting factor in step S105. In the next step S107, it is determined whether or not the highest score, that is, the score of the selected voice command candidate is larger than a predetermined threshold value. If the highest score is smaller than the predetermined threshold value, there is no voice command. After step S104, the operation returns to step S101 without performing a voice operation, and the voice recognition input device 10 is again in a state of waiting for voice input. At this time, the operator can be notified that the voice operation is not performed, for example, by using the synthesized voice, alarm, or display.

If the highest score is larger than the predetermined threshold, the process proceeds to step S108, and the voice command showing the highest score is determined as the result of the voice recognition process. The determined voice command is output to the command conversion unit 112, and the command conversion unit 112 uses the voice recognition data table to convert the determined voice command into an operation command corresponding to the voice command (S109).

In the next step S110, the command conversion unit 112 outputs the converted operation command to the command analysis unit 113 and the system operation determination unit 140. The command analysis unit 113 updates and generates the operation history including the input operation command and records it in the log collection DB 16. The system operation determination unit 140 outputs the input operation command to the image capturing device 20. In the image pickup device 20, an operation according to the input operation command is executed.

As described above, according to this embodiment, a plurality of voice commands are associated with an operation command and recorded, and a high weighting coefficient is given to each operation command in descending order of frequency of use, and the voice recognition process is used. It is possible to perform the voice recognition processing with high accuracy without depending on the habits and preferences of the utterance which differ for each operator. In addition, by recording the frequency of use of the operation command and updating the weighting coefficient according to the frequency of use, it is possible to improve the accuracy of the voice recognition process for the operation command that the operator frequently uses over time. The operator's voice operation instruction can be accurately recognized.

(Modification)
In the above-described first embodiment, the example in which the command analysis unit 113 updates the weighting factor based on the usage frequency of the operation command has been described. In this modification, an example in which the weighting factor is updated based on the input voice command will be described.

The command analysis unit 113 analyzes the operation history to calculate the detection frequency of each voice command within a certain period of time for the voice command input from the command conversion unit 112 before the operation command conversion. Then, the offset coefficient V for updating the weighting coefficient is calculated based on the calculated detection frequency of the voice command.

The offset coefficient V can be determined by the cumulative number of times the voice command code corresponding to each voice command is issued according to the graph shown in FIG. 5, for example. In addition, the offset coefficient V may be determined based on the ratio of each voice command to the total number of voice command codes issued in a certain period.

The command analysis unit 113 updates the weight coefficient by multiplying each calculated weight coefficient by the calculated offset coefficient V corresponding to the operation command. Also in this case, it is preferable that the command analysis unit 113 updates the weighting factor by subtracting from the number of times the voice command has been integrated when erroneous detection occurs in the voice recognition process of the voice recognition unit 111.

As described above, in this modified example, a voice command based on the utterance of the operator is given a high weighting coefficient in descending order of detection frequency, and the weighting coefficient is used in the voice recognition process, so that the utterance habits and preferences that are different for each operator The voice recognition process can be performed accurately without depending on. Further, by recording the number of times of detecting a voice command and updating the weighting coefficient according to the number of times of detection, it is possible to improve the accuracy of voice recognition for a voice command that is frequently detected by the operator over time. The accuracy of voice recognition can be further improved as compared with the first embodiment, and the operator's voice operation instruction can be accurately recognized.

It should be noted that it is also possible to update by multiplying the result obtained by multiplying the offset coefficient T based on the use frequency of the operation command and the offset coefficient V based on the detection frequency of the voice command, by multiplying the weight coefficient. In this case, the voice recognition accuracy of the frequently used operation command and the frequently detected voice command is further improved. Also in this case, when an erroneous detection occurs in the voice recognition process of the voice recognition unit 111, the command analysis unit 113 subtracts the weighting factor from the number of times the operation command has been used and the number of times the voice command has been integrated. It is preferable to update.

(Second embodiment)
In the above-described first embodiment and its modification, an example in which one voice recognition data table is stored in the voice recognition DB 15 has been described. In the present embodiment, a plurality of voice recognition data tables as shown in FIG. 2 are stored in the voice recognition DB 15 depending on the operation status of the external device to which the voice recognition input device 10 is applied. The data table used for the voice recognition processing is switched according to the operation status.

The operation status of the device may be, for example, before and after the start of inspection, the inspection type, the presence or absence of X-ray output, and the like, and the voice recognition DB has a plurality of voice recognition data tables tg001 to tg*** according to these situations in advance. Is stored. An example of the voice recognition data table is as shown in FIG. Further, in the voice recognition DB, data tables showing the operation status of the device as shown in FIG. 6 are stored, and an appropriate voice recognition data table is selected by referring to these data tables. 6A is a data table showing the state of the inspection start information, FIG. 6B is a data table showing the state of the inspection type, FIG. 6C is a data table showing the state of the X-ray irradiation information, and FIG. ) Is a data table showing the device operation status and the classification of the voice recognition database.

The flow of switching the voice recognition data tables tg001 to tg*** will be described below with reference to the flowchart of FIG.
When the voice recognition input device 10 is activated, the system operation determination unit 14 sequentially acquires the information related to the device operation status from the image capturing device 20 (step S201). In the present embodiment, the system operation determination unit 14 acquires, for example, inspection start information, inspection type information, and X-ray irradiation information. When the information related to the device operation status is acquired, the system operation determination unit 14 determines whether or not there is a change in each acquired information compared with the previous state (step S202).

In the determination in step S202, if at least one of the inspection start information, the inspection type information, and the X-ray irradiation information has changed, the process proceeds to step S203, and the apparatus operation status command St is generated. As shown in FIG. 6D, the device operation status command St is composed of three pieces of information, that is, inspection start information, inspection type, and X-ray irradiation information, and the voice recognition data table to be applied is determined according to these combinations. It is like this.

The system operation determination unit 14 outputs the generated device operation status command St to the voice recognition unit 111 (step S205), and the voice recognition unit 111 selects the voice recognition data table according to the input device operation status command St. , Switch. Each piece of information of the apparatus operation status command St, for example, the examination start information is ao1 indicating “after examination start”, the examination type is 1001 indicating “Abdomen (abdomen)”, and the X-ray irradiation information is c01 indicating “irradiating”. In some cases, the voice recognition database tg003 is selected.

Even when a plurality of voice recognition data tables are stored in the voice recognition DB as described above, the weighting factor can be updated as in the first embodiment and its modification.

When generating the operation history, the command analysis unit 113 records the operation command code indicating the operation command and the voice command code indicating the voice command, as well as the command code related to the presence or absence of the inspection start and the inspection type (see FIG. 6 ). .. By doing so, it is possible to update only the weighting coefficient of the voice recognition database that needs updating.

As described above, according to the present embodiment, it is possible to switch the voice recognition data table used for the voice recognition processing according to the operation status of the device to which the voice recognition input device is applied. Since each voice recognition data table can define a weighting coefficient that weights a frequently used operation and a frequently used voice command for each device operation situation, the accuracy of the voice recognition process can be improved.

10...Voice recognition input device, 11...CPU, 12...Voice input I/F, 13...Manual input I/F, 14...Memory, 15...Voice recognition DB, 16 ... log collection DB, 20... image pickup device, 30... display, 111... voice recognition unit, 112... command conversion unit, 113... command analysis unit, 120... voice Operation processing unit, 130... Manual operation processing unit, 140... System operation determination unit, 20... Image capturing device, 30... Display

Claims (7)

A voice recognition input device for inputting an operation command to an external device,
A storage unit that stores a voice recognition data table in which a plurality of voice commands are associated with one operation command and recorded, and a weighting coefficient corresponding to the frequency of use of the voice command is recorded for each voice command.
A voice recognition unit that receives a voice input, performs a voice recognition process with the voice input as a recognition target, and outputs a voice command corresponding to the voice input as a result of the voice recognition process;
A command conversion unit for converting the voice command into an operation command recorded corresponding to the voice command with reference to the voice recognition data table;
An operation determining unit that outputs the operation command to the external device,
The voice recognition unit selects a plurality of voice command candidates that may correspond to the voice input, and multiplies each of the plurality of voice command candidates by the weighting factor to determine the most probable voice command in the voice recognition process for the voice input. A voice recognition input device that outputs as a result. A command analysis unit is provided, which generates an operation history recording at least one of the operation command and the voice command corresponding to the operation command, and updates the weighting coefficient based on a result of analysis of the operation history. 1. The voice recognition input device according to 1. The voice recognition input device according to claim 2, wherein the command analysis unit updates the weighting factor based on at least one of a use frequency of the operation command within a certain period or an integrated number of voice commands. 4. The voice recognition input device according to claim 2, wherein the command analysis unit updates the weighting coefficient so as to decrease the weighting coefficient when an erroneous detection occurs in the voice recognition processing in the voice recognition unit. The storage unit stores a plurality of voice recognition data tables corresponding to the operating status of the external device,
The voice recognition input device according to any one of claims 1 to 4, wherein the voice recognition unit switches the voice recognition data table used for voice recognition processing according to an operation status of the external device. A voice recognition input program for causing a computer to input an operation command by voice to a medical image pickup device,
A plurality of voice commands are associated with one operation command and recorded, and a voice recognition data table in which a weighting coefficient corresponding to the frequency of use of the voice command is recorded for each voice command is referred to, and a voice input from the user is input. A voice recognition step of performing a voice recognition process as a recognition target and outputting a voice command corresponding to the voice input as a result of the voice recognition process;
A command conversion step of converting the voice command into an operation command recorded corresponding to the voice command with reference to the voice recognition data table;
An operation determining step of outputting the operation command to the medical image capturing device,
In the voice recognition step, a plurality of voice command candidates that may correspond to the voice input are selected, and the most probable voice command is multiplied by the weighting factor of each of the plurality of voice command candidates to perform the voice recognition process for the voice input. Speech recognition input program to output as a result. A voice recognition input device according to any one of claims 1 to 5,
The voice recognition input device according to any one of claims 1 to 5 as an external device, and the medical image pickup device according to any one of claims 1 to 5, wherein the voice recognition input device according to any one of claims 1 to 5. A medical image pickup system for giving an operation instruction to the medical image pickup apparatus by voice recognition input.

Images