Electronic device and operating method thereof

Abstract

The electronic device includes a memory and a processor, the memory storing one or more instructions, the processor being configured to execute the one or more instructions stored in the memory to receive audio data corresponding to user speech, identify user speech features based on the received audio data, determine parameters for performing voice activity detection through the identified user speech features, and perform voice activity detection on the received audio data for the user speech by using the determined parameters.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application is based on and claims the benefit of priority from Korean Patent Application No. 10-2019-0113010 filed on September 11, 2019, in the Korean Intellectual Property Office, the entire contents of which are incorporated herein by reference. Technical Field

[0003] Various embodiments of the present disclosure relate to an electronic device and an operating method thereof, and more particularly, to an electronic device and an operating method thereof for detecting voice activity detection regarding a user's speech. Background Art

[0004] Speech recognition refers to the process of converting an audio signal acquired by a sound sensor (such as a microphone) into text data (such as a word or sentence). As the number of devices and services for providing speech recognition increases, speech recognition has been used in various places and in various forms.

[0005] The speech recognition system performs voice activity detection (VAD) before performing speech recognition to distinguish between speech parts and non-speech parts in the audio signal. For example, since it is not necessary to perform speech recognition on the non-speech parts of the audio signal where the user has not spoken, speech recognition is performed only on the speech parts detected by VAD.

[0006] The speech recognition system performs VAD by applying an average parameter value obtained as a result of analyzing the speech of a person in general. In this case, a VAD error may occur when the speech rate of the user is relatively low or high.

[0007] Public Content

[0008] Various embodiments of the present disclosure provide an electronic device capable of changing parameters according to user utterance characteristics to perform voice activity detection and an operating method thereof.

[0009] Additional aspects will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the embodiments presented in the disclosure.

[0010] According to an embodiment of the present disclosure, an electronic device includes a memory storing one or more instructions and a processor, the processor being configured to execute the one or more instructions stored in the memory to receive audio data corresponding to user speech, to identify user speech features based on the received audio data, to determine parameters for performing voice activity detection by using the identified user speech features, and to perform voice activity detection on the received audio data regarding the user speech by using the determined parameters.

[0011] In one embodiment of the present disclosure, the audio data may include a speech portion corresponding to a trigger word uttered by the user.

[0012] In one embodiment of the present disclosure, determining the parameter includes: comparing the user's speech feature with a reference speech feature regarding the trigger word, and changing the parameter based on a result of the comparison.

[0013] In one embodiment of the present disclosure, identifying the user's speech features includes: when the audio data includes a speech portion corresponding to a trigger word, identifying the user's speech features based on the received audio data.

[0014] In one embodiment of the present disclosure, performing voice activity detection may include: when a user speaks, detecting the starting point of the user's speech in the received audio data and the ending point of the user's speech in the received audio data to distinguish between the speech portion of the received audio data and the non-speech portion of the received audio data.

[0015] In one embodiment of the present disclosure, the recognized speech feature of the user may include at least one of the user's speech rate, speech energy level, or pronunciation voice height.

[0016] In one embodiment of the present disclosure, the parameters for performing voice activity detection may include: one or more of an energy threshold, a delay time, or an automatic end time for identifying a speech portion when the user is uttering on the received audio data.

[0017] In one embodiment of the present disclosure, the electronic device may further include a communicator configured to receive audio data corresponding to the user's utterance.

[0018] In one embodiment of the present disclosure, the electronic device may further include a microphone configured to receive a user speech and convert the user speech into audio data.

[0019] According to another embodiment of the present disclosure, an operating method of an electronic device includes: receiving audio data corresponding to user speech; identifying user speech features based on the received audio data; determining parameters for performing voice activity detection by using the identified user speech features; and performing voice activity detection on the received audio data regarding the user speech by using the determined parameters.

[0020] Identifying the user's speech features includes: when the audio data includes a speech portion corresponding to a trigger word uttered by the user, identifying the user's speech features based on the received audio data.

[0021] Determining parameters for performing voice activity detection includes comparing the recognized user utterance features with reference utterance features regarding the trigger word.

[0022] Identifying the user's speech features includes: when the audio data includes a speech portion corresponding to the trigger word, identifying the user's speech features based on the audio data.

[0023] Performing voice activity detection includes: in the case of a user speaking, detecting a start point of the user's speech and an end point of the user's speech to distinguish between a speech portion of the received audio data and a non-speech portion of the received audio data.

[0024] Determining the parameter includes determining the parameter for performing voice activity detection based on at least one of a speech rate, a speech energy level, or a pronunciation voice height of the user.

[0025] The parameters for performing voice activity detection include at least one of an energy threshold for identifying a speech portion in case of a user uttering a voice, a delay time, or an automatic end time.

[0026] Receiving the audio data of the user's speech includes: receiving the audio data corresponding to the user's speech through the communicator.

[0027] Receiving audio data of user speech includes: receiving the user speech through a microphone, and converting the user speech into audio data.

[0028] A non-transitory computer-readable recording medium having recorded thereon a program for executing the operating method according to an embodiment of the present application on a computer.

[0029] The processor is further configured to perform voice activity detection based on the changed user utterance characteristics.Performing voice activity detection also includes performing voice activity detection based on the changed user utterance characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The above and other aspects, features and advantages of certain embodiments of the present disclosure will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:

[0031] Figure 1 is a schematic diagram showing an electronic device according to an embodiment of the present disclosure;

[0032] Figure 2A is a schematic diagram for describing an embodiment in which an error occurs when performing voice activity detection on audio data of a user's speech;

[0033] Figure 2B is a schematic diagram for describing another embodiment in which an error occurs when performing voice activity detection on audio data of a user's speech;

[0034] Figure 3 is a block diagram of an electronic device according to an embodiment of the present disclosure;

[0035] Figure 4A and Figure 4B is a schematic diagram for describing a method for analyzing audio data corresponding to a trigger word according to an embodiment of the present disclosure;

[0036] Figure 5 is a flow chart of an operating method of an electronic device according to an embodiment of the present disclosure;

[0037] Figure 6 is a flowchart of an operating method of an electronic device according to another embodiment of the present disclosure;

[0038] Figure 7 is a schematic diagram for describing a method for changing parameters for performing voice activity detection based on user speech features, performed by an electronic device according to an embodiment of the present disclosure;

[0039] Figure 8 is a schematic diagram for describing a method for changing parameters for performing voice activity detection by using a database including user information, performed by an electronic device according to an embodiment of the present disclosure;

[0040] Fig. 9 is a schematic diagram showing a speech recognition system according to an embodiment of the present disclosure;

[0041] Fig.10 is a block diagram of an electronic device according to an embodiment of the present disclosure; and

[0042] Fig.11 is a block diagram of an electronic device according to another embodiment of the present disclosure. DETAILED DESCRIPTION

[0043] The terms used in the specification will be briefly described, and then the present disclosure will be described in detail.

[0044] In the present disclosure, considering the functions of the present disclosure, currently widely used common terms are selected as much as possible, but non-common terms may also be selected according to the intentions of those skilled in the art, precedents or new technologies, etc. Some terms may be arbitrarily selected by the applicant of the present application. In this case, the meanings of these terms will be explained in detail in the corresponding parts of the present disclosure. Therefore, the terms used in this article should not be defined based on their names, but should be defined based on their meanings and the entire context of the present disclosure.

[0045] It should be understood that when an element is referred to as "comprising" another element, unless otherwise specified, the element may further include other elements. When used herein, terms such as "unit", "module" and the like represent a unit for processing at least one function or operation, which can be implemented by hardware, software or a combination of hardware and software.

[0046] In an embodiment of the present disclosure, the term “user” refers to a viewer who views an image displayed on an electronic device or a person who controls a function or operation of the electronic device, and may include an administrator or an installation engineer.

[0047] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement the embodiments of the present disclosure. However, the present disclosure can be presented in a variety of different forms and is not limited to the embodiments of the present disclosure set forth herein. For clarity, parts not related to the explanation of the present disclosure are omitted from the accompanying drawings, and the same parts are represented by the same reference numerals throughout the specification.

[0048] Figure 1 is a schematic diagram showing an electronic device according to an embodiment of the present disclosure.

[0049] refer to Figure 1 , the electronic device 100 according to the embodiment of the present disclosure may be embodied as various types of electronic devices, such as a television, a set-top box, a mobile phone, a speaker, a tablet computer, a digital camera, a video camera, a laptop computer, a desktop computer, an e-book terminal, a digital broadcast terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, an MP3 player, an Internet Protocol television (IPTV), a digital television (DTV), a wearable device, etc. Alternatively, the electronic device 100 according to the embodiment of the present disclosure may be a household appliance such as a refrigerator, a washing machine, an air conditioner, a vacuum cleaner, a light bulb, or an air purifier. Alternatively, the electronic device 100 may be a fixed electronic device set at a fixed position, or a mobile electronic device in a movable form, or may be a digital broadcast receiver capable of receiving digital broadcast services.

[0050] In one embodiment of the present disclosure, the control device 200 may be implemented as various types of devices for controlling the electronic device 100, such as a remote controller or a mobile phone. An application for controlling the electronic device 100 may be installed in the control device 200, and the control device 200 may control the electronic device 100 by using the installed application. However, the embodiments of the present disclosure are not limited thereto. In addition, the control device 200 may control the electronic device 100 by using infrared (IR), Bluetooth (BT), Wi-Fi, etc.

[0051] The user may speak or utter a voice to the electronic device. The user utterance may include natural language that instructs the electronic device 100 to perform a certain function (eg, control the operation of hardware / software components included in the electronic device 100, content search, etc.).

[0052] In one embodiment of the present disclosure, the electronic device 100 may use an internal or external audio input module (eg, a microphone) to convert the user's speech into an audio signal (audio data), which is an electrical signal.

[0053] The user can speak to the control device 200, and the control device 200 can convert the user's speech into an audio signal (audio data), which is an electrical signal, using an internal or external microphone. The control device 200 can send the audio signal to the electronic device 100, and the electronic device 100 can receive the audio signal corresponding to the user's speech from the control device 200 through the communication module.

[0054] In one embodiment of the present disclosure, when a trigger word (or wake-up word) is recognized or a predetermined input is received, the electronic device 100 may start voice recognition of an audio signal. The predetermined input may be an input for pressing a physical button included in the electronic device 100, or an input for selecting a graphical interface (e.g., an icon, etc.) displayed on the electronic device 100 or on the display of the electronic device 100. Alternatively, the predetermined input may be an input for selecting a physical button included in a control device 200 that communicates with the electronic device 100, or an input for selecting a graphical interface displayed on a display of an external device.

[0055] When voice recognition is started for an audio signal, the electronic device 100 according to an embodiment of the present disclosure can perform voice activity detection by analyzing the audio signal of the user's speech. Voice activity detection refers to the process of distinguishing the voice part and the silent part of the audio signal from the user's speech when the user speaks by detecting the beginning and end of the speech.

[0056] Accordingly, the audio signal or audio data corresponding to the silent portion is not processed, thereby minimizing resource usage of the electronic device 100. Alternatively, when voice recognition is performed on the audio signal or audio data through an external server or external device, the audio signal or audio data corresponding to the silent portion may not be sent to the external server or external device to reduce network traffic.

[0057] In one embodiment of the present disclosure, the electronic device 100 may divide the audio data of the user's speech into multiple frames, and identify whether each of the multiple frames is a speech part or a non-speech part. For example, the electronic device 100 may identify whether each of the multiple frames is a speech part or a non-speech part based on a speech energy threshold, a delay time, an automatic end time, etc. When the speech energy level corresponding to the frame is greater than or equal to the threshold, the electronic device 100 may identify that the frame corresponds to the speech part, and when the speech energy level corresponding to the frame is less than the threshold, it may be identified that the frame corresponds to the non-speech part. Alternatively, when the non-speech part lasts for a delay time or longer, the electronic device 100 may determine that the speech has ended. When the end point of the speech is not detected, the electronic device 100 may determine that the speech has ended when the automatic end time has passed. However, the embodiments of the present disclosure are not limited to this.

[0058] Figure 2A is a schematic diagram for describing an embodiment in which an error occurs when performing voice activity detection on audio data of a user's utterance, and Figure 2B is a schematic diagram for describing another embodiment in which an error occurs when voice activity detection is performed on audio data of a user's utterance.

[0059] Figure 2A The audio data 210 is shown when a user having a relatively slow speaking rate speaks.

[0060] When the set delay time has passed, the voice activity detector determines that the speech has ended. When the user's speech rate is low, the time period between one word and another (the time period after one word is spoken until the next word is spoken) becomes longer than the set delay time, so it can be determined that the speech has ended even if the speech has not ended. For example, if Figure 2A As shown, the user's actual speech start point is P1, the actual speech end point is P2, and the actual voice part (speech part) is T1, but the voice activity detector may determine that the user's speech start point is P3, the speech end point is P4, and the user's voice part (speech part) is T2. Therefore, the user's speech may be lost.

[0061] Alternatively, the voice activity detector determines that speech has ended when the set automatic end time has passed. When the user's speech rate is high, it may be determined that speech has ended after a long time, even if the speech has ended long ago. For example, Figure 2B As shown in the figure, the actual speech start point of the user is P5, the actual speech end point is P6, and the actual voice part (speech part) is T3, but the voice activity detector may determine that the user's speech start point is P7, the speech end point is P8, and the user's voice part (speech part) is T4. Therefore, unnecessary waiting time may occur.

[0062] When voice activity detection is performed by applying the same parameters without considering the characteristics of the user's utterance, the accuracy and efficiency of the voice activity detection may be low.

[0063] Figure 3 is a block diagram of an electronic device according to an embodiment of the present disclosure.

[0064] Reference Figure 3 , the electronic device 100 according to an embodiment of the present disclosure may include a voice inputter 310 , a voice activity detector (VAD) 320 , a trigger word recognizer 330 , and a trigger word analyzer 340 .

[0065] The voice inputter 310 may receive audio data of a user's speech. For example, when the electronic device 100 receives the user's speech through an internal or external audio input module (e.g., a microphone), the electronic device 100 may receive audio data of the user's speech from the internal or external audio input module. Alternatively, when the user's speech is received using an internal or external microphone of an external device, the voice inputter 310 may receive audio data from the external device through the communication module.

[0066] The voice inputter 310 may perform pre-processing to reduce or remove noise in the audio data.

[0067] The voice activity detector 320 can detect the start and end points of the user's utterance by dividing the audio data of the user's utterance into a speech part (speech part) including the user's voice and a non-speech part (non-speech part or silent part) not including the user's voice. The voice activity detector 320 can divide the audio data into a plurality of frames and identify whether each of the plurality of frames corresponds to a speech part or a non-speech part.

[0068] Based on the parameters set in the voice activity detector 320, the voice activity detector 320 can distinguish between a voice part and a non-voice part. For example, the voice activity detector 320 can identify whether each of the plurality of frames is a voice part or a non-voice part based on parameters such as a speech energy threshold, a delay time, an automatic end time, etc. The voice activity detector 320 can identify the frame as a voice part when the speech energy level corresponding to the frame is greater than or equal to the threshold, and identify the frame as a non-voice part when the speech energy level corresponding to the frame is less than the threshold.

[0069] Alternatively, when the non-speech portion continues for the delay time or longer, the voice activity detector 320 may determine that the speech has ended. When the speech end point is not detected, the voice activity detector 320 may determine that the speech has ended when the automatic end time has passed.

[0070] Additionally, the voice activity detector 320 may perform a process of removing non-speech portions from the user's audio data.

[0071] The trigger word identifier 330 may identify whether the user's speech corresponds to a trigger word. For example, the trigger word identifier 330 may identify whether the speech portion of the audio data detected by the VAD 320 corresponds to a trigger word. The trigger word is a word indicating the start of speech, and may be referred to as a wake-up word, a speech start word, etc., but is not limited thereto.

[0072] The trigger word may be set in advance, and the trigger word identifier 330 may include an isolated word identifier that searches for a word with a maximum probability, a keyword search engine that searches for a preset keyword of the audio data.

[0073] Alternatively, the trigger word recognizer 330 may perform a process of acquiring a speech portion of the audio data in the form of text or words, and recognize whether the text or words acquired as a result of the processing correspond to a trigger word.

[0074] The trigger word analyzer 340 may analyze the user speech corresponding to the trigger word to identify the user speech characteristics. The trigger word analyzer 340 may analyze the speech energy level and rate of the speech part corresponding to the trigger word to identify the user speech characteristics. The trigger word analyzer 340 may change the parameter settings of the VAD 320 based on the user speech characteristics.

[0075] For example, the trigger word analyzer 340 may compare the speech rate (or speech time) of the user who issued the trigger word with the reference speech rate (or reference speech time) to change the delay time for voice activity detection. In this case, the reference speech rate is a preset value and can be an average value of the speech rate of an average person.

[0076] Alternatively, the speech energy level of the user who issued the trigger word can be compared with the reference speech energy level to change the speech energy threshold for voice activity detection. In this case, the reference speech energy level can be the average of the speech energy levels of various people.

[0077] The trigger word analyzer 340 may calculate the values ​​Vc and Ec by using the following Formula 1.

[0078] Formula 1

[0079]

[0080] Here, Vc represents the reference speech time t avg (Statistically calculated average time during which the trigger word is issued) and the time t when the user issues the trigger word trigger Ec represents the energy level of the reference speech. avg (Statistically calculated average speech energy level of the trigger word) and the speech energy level of the user who issued the trigger word e trigger The ratio between .

[0081] The trigger word analyzer 340 may change parameters such as a delay time, an utterance energy threshold, etc. of the VAD 320 based on the values ​​Vc and Ec.

[0082] For example, the trigger word analyzer 340 can change the delay time by using the Vc value. When the user's utterance rate is low, the Vc value decreases, and as the Vc value decreases, the delay time can be changed to increase. On the other hand, when the user's utterance rate is high, the Vc value increases, and as the Vc value increases, the delay time can be changed to decrease. Figure 4A and Figure 4B Describe in detail.

[0083] In addition, the trigger word analyzer 340 can change the speech energy threshold by using the Ec value. When the user speech energy level is low, the Ec value increases, and as the Ec value increases, the speech energy threshold can be changed to decrease. On the other hand, when the user speech energy level is high, the Ec value decreases, and as the Ec value decreases, the speech energy threshold can be changed to increase. Therefore, when a user with a low speech energy level speaks, a lower speech energy threshold is lowered to prevent the user's speech from being recognized as noise, and when a user with a high speech energy level speaks, a higher speech energy threshold is increased to prevent the user's speech from being recognized as noise.

[0084] The trigger word analyzer 340 may variably set the size of a window (e.g., a Hamming window) for analyzing a user speech feature (e.g., a user's speech rate or speech energy level) based on the user speech feature. For example, the size of the window may be reduced when the user's speech rate is high, and the size of the window may be increased when the user's speech rate is low, to analyze the user speech feature.

[0085] Furthermore, the trigger word analyzer 340 may change parameter settings of the VAD 320 , a signal-to-noise ratio, a zero-crossing rate Z, a result of spectrum analysis, etc. based on the voice pitch of the user who utters the trigger word.

[0086] Figure 4A and Figure 4B It is a schematic diagram for describing a method for analyzing audio data corresponding to a trigger word according to an embodiment of the present disclosure.

[0087] Figure 4A and Figure 4B The audio data corresponding to the trigger word when different users utter the trigger word is shown.

[0088] refer to Figure 4A and Figure 4B , the first audio data 410 represents the audio data corresponding to the trigger word when the first user issues the trigger word, and the second audio data 420 represents the audio data corresponding to the trigger word when the second user issues the trigger word issued by the first user.

[0089] The length of the portion of the utterance in which the trigger word is spoken may vary from user to user, and when voice activity detection is performed using the same parameters, the accuracy and efficiency of voice activity detection may be different from those described above. Figure 2A and Figure 2B As described low.

[0090] In one embodiment of the present disclosure, the trigger word analyzer 340 may calculate the first utterance time T1 (=Te1 −Ts1 ) by detecting the utterance start point Ts1 and the utterance end point Te1 from the first audio data 410 .

[0091] The trigger word analyzer 340 may calculate a ratio Vc1 between a reference speech time T0 of uttering the trigger word and a first speech time T1 of the first user, and based on the ratio Vc1, change the delay time of the VAD 320. In this case, the reference speech time T0 may be a speech time when a normal person utters the trigger word.

[0092] For example, when the delay time of the VAD 320 is set to the average delay time Th, the delay time may be changed to Th / Vc1.

[0093] Therefore, when the speech rate of the first user is higher than the reference speech rate, the ratio Vc1 is greater than 1 and the delay time is reduced, so the time required to determine the speech end point Te1 can be reduced. Therefore, when a user with a high speech rate speaks, unnecessary waiting time can be reduced.

[0094] In addition, the trigger word analyzer 340 may calculate the second utterance time T2 (=Te2 − Ts1 ) by detecting the utterance start point Ts1 and the utterance end point Te2 in the second audio data 420 .

[0095] The trigger word analyzer 340 may calculate a ratio Vc2 between the reference speech time T0 and the second speech time T2 of the second user, and change the delay time of the VAD 320 based on the ratio Vc2. For example, when the delay time of the VAD 320 is set to the average delay time Th, the delay time may be changed to Th / Vc2.

[0096] Therefore, when the speech rate of the second user is lower than the reference speech rate, the ratio Vc2 is less than 1 and the delay time increases, so the time required to determine the speech end point Te2 can be increased. Therefore, when the user with a low speech rate speaks, even when the silent part is relatively long, the silent part between words can not be determined as the speech end point. Therefore, when the user with a low speech rate speaks, the loss of speech can be prevented.

[0097] Figure 5 is a flowchart of an operating method of an electronic device according to an embodiment of the present disclosure.

[0098] Reference Figure 5 According to an embodiment of the present disclosure, the electronic device 100 may obtain audio data of a user's speech (S510). The electronic device 100 may obtain audio data by receiving the user's speech through an audio input module, or obtain audio data of the user's speech from an external device through a communication module. However, the embodiments of the present disclosure are not limited thereto.

[0099] Furthermore, when a trigger word is recognized or a predetermined input is received, the electronic device 100 may start voice recognition of the audio data.

[0100] In one embodiment of the present disclosure, when voice recognition is started on an audio signal, the electronic device 100 may perform voice activity detection by analyzing the audio signal of the user's speech.

[0101] The electronic device 100 may recognize a user utterance feature based on a voice portion detected from audio data of the user utterance (S520).

[0102] User speech features may include, but are not limited to, the user's speech energy, speech rate, speech time, speech pitch, etc.

[0103] For example, when a user speaks a preset keyword, the electronic device 100 can analyze the user's speech characteristics by measuring the user's speech energy, speech rate, speech time, voice pitch, etc., and comparing them with reference values ​​of the preset keyword (for example, average speech energy level, average speech rate, average speech time, average voice pitch, etc.).

[0104] In this case, the reference value of the preset keyword can be pre-stored in the electronic device 100 or obtained from an external device. The preset keyword according to an embodiment of the present disclosure may be, but is not limited to, a trigger word, a wake-up word, a speech start word, or a similar word indicating the start of speech.

[0105] The electronic device 100 may change parameters for voice activity detection based on the user utterance characteristics (S530).

[0106] The electronic device 100 may change the parameters used to perform voice activity detection (e.g., speech energy threshold, speech rate threshold, voice pitch threshold, delay time, automatic end time, window size for speech analysis, etc.) based on the user's speech energy level, speech rate, speech time, and voice pitch.

[0107] The electronic device 100 may perform voice activity detection on the audio data of the user's utterance by using the changed parameters (S540).

[0108] Figure 6 is a flowchart of an operating method of an electronic device according to another embodiment of the present disclosure.

[0109] Reference Figure 6 , the electronic device 100 according to an embodiment of the present disclosure may receive audio data of a user's speech ( S610 ).

[0110] For example, when the electronic device 100 receives user speech through an internal or external audio input module (e.g., a microphone), the electronic device 100 may receive audio data of the user speech from the internal or external audio input module. Alternatively, when receiving user speech using an internal or external microphone of an external device, the electronic device 100 may receive audio data from the external device through a communication module (communicator).

[0111] The electronic device 100 may perform voice activity detection on the audio data (S620).

[0112] The electronic device 100 can distinguish between a speech part (speech part) and a non-speech part (non-speech part) from the audio data based on preset parameters. For example, the electronic device 100 can identify whether each of the multiple frames included in the audio data corresponds to a speech part or a non-speech part based on parameters such as a speech energy threshold, a delay time, and an automatic end time. When the speech energy level corresponding to the frame is greater than or equal to the threshold, the electronic device 100 can identify the frame as a speech part, and when the speech energy level corresponding to the frame is less than the threshold, the frame can be identified as a non-speech part.

[0113] Alternatively, when the non-voice portion continues for the delay time or longer, the electronic device 100 may determine that the speech has ended. When the speech end point is not detected, the electronic device 100 may determine that the speech has ended when the automatic end time has passed.

[0114] The electronic device 100 may identify whether the audio data of the speech part recognized as the voice part corresponds to a trigger word (S630). The trigger word is a word indicating the start of speech and may be called a wake-up word, a speech start word, etc., but is not limited thereto.

[0115] When the audio data is identified as a trigger word, the electronic device 100 may analyze the user speech corresponding to the trigger word to identify the user speech feature (S640). For example, the electronic device 100 may analyze the user speech feature by comparing the speech energy level, speech rate, speech time, and voice pitch of the user who issued the trigger word with a reference value of the trigger word (for example, the average speech energy level, average speech rate, average speech time, and average voice pitch when an average person issues the trigger word).

[0116] The electronic device 100 may change the settings of the voice activity detector based on the user's speech characteristics (S650). For example, the electronic device 100 may change the parameters used to perform voice activity detection (e.g., speech energy threshold, speech speed threshold, speech pitch threshold, delay time, automatic end time, window size of speech analysis, etc.) based on the user's speech energy level, speech rate, speech time, and voice pitch.

[0117] When the user's free utterance is input after receiving the trigger word ( S660 ), the electronic device 100 may perform voice activity detection on the user's free utterance by using the changed parameter ( S670 ).

[0118] Figure 7 is a schematic diagram for describing a method for changing parameters for performing voice activity detection based on user utterance features, which is performed by an electronic device according to an embodiment of the present disclosure.

[0119] refer to Figure 7 According to an embodiment of the present disclosure, the electronic device 100 may receive specific keywords to determine user speech features.

[0120] For example, when the user issues a trigger word or performs a predetermined input, the electronic device 100 displays a message 710 indicating that a keyword (e.g., "Hello ABC") is spoken on its display. The predetermined input may be an input for pressing a physical button included in the electronic device 100 or in the control device 200 that controls the electronic device 100, or may be an input for selecting a graphical interface (e.g., an icon) displayed on the display of the electronic device 100 or on the display of the control device 200.

[0121] The keyword is a keyword preset in the electronic device 100 and is a combination of simple words rather than a long sentence. Alternatively, the electronic device 100 may store reference speech features corresponding to the keyword (e.g., speech energy level, speech rate, speech time, voice pitch, etc.). In this case, the reference speech features corresponding to the keyword may refer to the average speech energy, average speech rate, average speech time, average voice pitch, etc. of an average person when uttering the keyword.

[0122] The user may issue a keyword (e.g., "Hello ABC") in response to the displayed message. The user may speak to the electronic device 100 or the control device 200. When the user speaks to the electronic device 100, the electronic device 100 may receive audio data of the user's speech through the audio input module. When the user speaks to the control device 200, the electronic device 100 may receive audio data of the user's speech through the communication module.

[0123] The electronic device 100 can detect the speech part by performing voice activity detection on the received audio data, and identify whether the keyword is included in the speech part to obtain the user speech feature for the keyword. For example, the electronic device 100 can obtain the speech feature (e.g., speech energy level, speech rate, speech time, voice pitch, etc.) of the user who uttered the keyword.

[0124] The electronic device 100 can compare the acquired user speech features with the reference speech features to change the parameters used to perform voice activity detection (for example, speech energy threshold, speech speed threshold, voice pitch threshold, delay time, automatic end time, window size for speech analysis, etc.).

[0125] The electronic device 100 may use the changed parameters to perform voice activity detection when the user speaks. For example, the electronic device 100 may use the changed parameters to perform voice activity detection on free speech input by the user after uttering a keyword. However, the embodiments of the present disclosure are not limited thereto.

[0126] Figure 8 is a schematic diagram for describing a method of changing parameters for performing voice activity detection by using a database including user information, which is performed by an electronic device according to an embodiment of the present disclosure.

[0127] Reference Figure 8 According to an embodiment of the present disclosure, the electronic device 100 may include a database 800 for storing user information. According to an embodiment of the present disclosure, the database 800 may include: at least one of user facial information, user fingerprint information, voice activity detection parameters, or trigger word speech features (such as trigger word speech energy level, trigger word speech rate, etc.) corresponding to user identification information.

[0128] For example, when the first user logs in with his or her account information and utters a trigger word, the electronic device 100 may store the trigger word speech energy level and the trigger word speech rate of the first user in the database 800 as the first user information.

[0129] Alternatively, when the first user issues a trigger word, the electronic device 100 may obtain facial information of the first user by taking a photo of the first user, and store the obtained facial information of the first user as first user information in the database 800 .

[0130] Alternatively, the electronic device 100 may receive the fingerprint information of the first user through an external device or the like when the first user issues a trigger word, and store the received fingerprint information of the first user as the first user information in the database 800 .

[0131] Alternatively, the electronic device 100 may determine the first voice activity detection parameter based on the trigger word speech energy level and the trigger word speech rate of the first user, and store the determined first voice activity detection parameter as the first user information in the database 800.

[0132] The electronic device 100 may also store the user information of the second to fourth users in the database 800 in the same manner.

[0133] When the first user speaks, the electronic device 100 can perform facial recognition by taking a photo of the first user or perform fingerprint recognition by receiving a fingerprint image of the first user to identify the first user. In this case, even when the first user does not utter a trigger word, the parameters for performing voice activity detection can be changed based on the trigger word speech energy level, trigger word speech rate, and first voice activity detection parameter of the first user included in the first user information.

[0134] In addition, the electronic device 100 may store not only speech features about the trigger word but also speech features of the first user's previous speech input in the database 800 and change parameters for performing voice activity detection based on the speech features of the first user's previous speech input.

[0135] Fig. 9 is a schematic diagram showing a speech recognition system according to an embodiment of the present disclosure.

[0136] Reference Fig. 9 , a speech recognition system according to an embodiment of the present disclosure may include an electronic device 100 and a server 900. The server 900 may be connected to the electronic device 100 via a network or short-range communication. In one embodiment of the present disclosure, the server 900 may be a server that performs speech recognition.

[0137] In one embodiment of the present disclosure, the electronic device 100 may receive audio data of a user's speech. For example, the electronic device 100 may convert the user's speech (analog signal) into an audio signal (digital signal) as an electrical signal through an internal or external audio input module (such as a microphone), and send the audio signal to the processor.

[0138] Alternatively, when the user speaks to the control device 200 according to an embodiment of the present disclosure, the control device 200 can convert the user's speech (analog signal) into an audio signal (digital signal) through an internal or external microphone, and the audio signal is an electrical signal. The control device 200 can send the audio signal to the electronic device 100 through a communication module such as a Bluetooth (BT) module.

[0139] The electronic device 100 may detect a voice section by performing voice activity detection on an audio signal (audio data), and transmit the detected voice section to the server 900 .

[0140] In one embodiment of the present disclosure, the server 900 may perform speech recognition on the speech portion received from the electronic device 100. Speech recognition may be a process of obtaining text data corresponding to an audio signal. Speech recognition may include a speech-to-text (STT) process. For example, speech recognition may include a process of recognizing a speech signal emitted by a user as a character string. The text obtained as a result of speech recognition may be in the form of a sentence, word, or phrase in a natural language. However, embodiments of the present disclosure are not limited thereto.

[0141] The server 900 may identify whether a trigger word is included in the voice part through voice recognition.

[0142] When a trigger word is included in the speech portion, the server 900 may analyze the user speech corresponding to the trigger word to identify the user speech characteristics. For example, the server 900 may analyze the user speech characteristics by comparing the speech energy level, speech rate, speech time, and voice pitch of the user who issued the trigger word with a reference value of the trigger word (e.g., the average speech energy level, average speech rate, average speech time, and average voice pitch when an average person issues the trigger word).

[0143] The server 900 may transmit the result of analyzing the speech feature to the electronic device 100 .

[0144] Alternatively, when the server 900 identifies whether a trigger word is included in the speech portion and notifies the electronic device 100 of the result of the identification, the electronic device 100 may analyze the user speech corresponding to the trigger word to identify the user speech feature. The method of identifying the user speech feature by analyzing the user speech corresponding to the trigger word performed by the electronic device 100 has been described above with reference to Figures 3 to 8 It has been described in detail, so its detailed description is omitted here.

[0145] The electronic device 100 may change parameters for performing voice activity detection based on user utterance characteristics, and when a user's free utterance is input after a trigger word is input, voice activity detection may be performed on the user's free utterance by using the changed parameters.

[0146] The electronic device 100 may send the voice part detected from the user's free speech to the server 900, and the server 900 may perform voice recognition on the user's free speech and perform a certain function based on the result of the voice recognition. Alternatively, the result of the voice recognition may be sent to the electronic device 100 or another server (not shown).

[0147] When the electronic device 100 receives the speech recognition result, the electronic device 100 may execute a certain function corresponding to the speech recognition result. Alternatively, when another server receives the speech recognition result, the other server may execute a certain function based on the speech recognition result or control another electronic device to execute the function.

[0148] Alternatively, the electronic device 100 may perform voice recognition on the voice portion detected from the free speech, and perform a certain function based on the result of the voice recognition.

[0149] Fig.10 is a block diagram of an electronic device according to an embodiment of the present disclosure.

[0150] Reference Fig.10 , the electronic device 100 according to an embodiment of the present disclosure may include a microphone 110 , a processor 120 , a memory 130 , and a communicator 140 .

[0151] The microphone 110 according to an embodiment of the present disclosure may receive a sound signal from an external device or a speaker (a user of the electronic device 100). In one embodiment of the present disclosure, the microphone 110 may receive a voice of a user's speech. The microphone 110 may receive an external sound signal and convert the external sound signal into an electrical signal (audio data). The microphone 110 may use various noise removal algorithms to remove noise generated during the reception of the external sound signal.

[0152] In one embodiment of the present disclosure, the communicator 140 may send data or signals to an external device (e.g., a control device, a server, etc.), or receive data or signals from an external device under the control of the processor 120. The communicator 140 may send and receive data or signals according to at least one of a wireless LAN (e.g., Wi-Fi), Bluetooth, wired Ethernet, infrared (IR), low-power Bluetooth (BLE), ultrasound, or ZigBee to correspond to the performance and structure of the electronic device 100. Here, the communicator 140 may include at least one communication module that can send and receive data according to a communication standard corresponding to a wireless LAN (e.g., Wi-Fi), Bluetooth, wired Ethernet, IR, BLE, ultrasound, and ZigBee.

[0153] In one embodiment of the present disclosure, the communicator 140 may send data or signals to the control device or server and receive data or signals from the control device or server by using the same communication module (e.g., Wi-Fi module). Alternatively, in one embodiment of the present disclosure, the module for communicating with the control device (e.g., Bluetooth module or IR module) and the module for communicating with the server (e.g., Ethernet module or Wi-Fi module) may be different from each other.

[0154] In one embodiment of the present disclosure, the processor 120 may control the overall operation of the electronic device 100. In addition, the processor 120 may control other components included in the electronic device 100 to perform certain operations.

[0155] In one embodiment of the present disclosure, the processor 120 may execute one or more programs stored in the memory 130. The processor 120 may include a single core, a dual core, a triple core, a quad core, and multiples thereof. The processor 120 may include a plurality of processors.

[0156] In one embodiment of the present disclosure, the memory 130 may store various data, programs or applications for driving and controlling the electronic device 100. In addition, in one embodiment of the present disclosure, the memory 130 may store the above-mentioned Figure 8 For example, the memory 130 may store at least one of user facial information, user fingerprint information, voice activity detection parameters, or trigger word speech features corresponding to the user identification information as the user information.

[0157] The program stored in the memory 130 may include one or more instructions. The program (one or more instructions) or application program stored in the memory 130 may be executed by the processor 120.

[0158] In one embodiment of the present disclosure, the processor 120 may acquire audio data of a user's speech and perform voice activity detection on the acquired audio data. The processor 120 may perform voice activity detection to detect a speech portion (speech portion) in the audio data and identify user speech features based on the detected speech portion. For example, the processor 120 may identify whether the detected speech portion corresponds to a preset keyword or trigger word. The processor 120 may analyze the user speech corresponding to the preset keyword or trigger word to identify the user speech features.

[0159] The processor 120 may change the parameters for performing voice activity detection based on user speech characteristics (e.g., speech energy level, speech rate, speech time, voice pitch, etc.). For example, the processor 120 may change the speech energy threshold, speech rate threshold, voice pitch threshold, delay time, automatic end time, window size for analyzing speech, etc. based on the user speech characteristics.

[0160] In addition, the processor 120 may also identify the user who is speaking, and identify user utterance features or change parameters for performing voice activity detection by using user information corresponding to the identified user.

[0161] The processor 120 may perform voice activity detection on the user's utterance by using the changed parameters.

[0162] Fig.11is a block diagram of an electronic device 1000 according to another embodiment of the present disclosure.

[0163] Fig.11 The electronic device 1000 in the embodiment can be used in conjunction with the above-mentioned reference Figures 1 to 10 The described electronic device 100 according to the embodiment of the present disclosure corresponds.

[0164] Reference Fig.11 According to an embodiment of the present disclosure, the electronic device 1000 includes: a tuner 1040, a processor 1010, a display 1020, a communicator 1050, a detector 1030, an inputter / outputter 1070, a video processor 1080, an audio processor 1085, an audio outputter 1060, a memory 1090 and a power supply 1095.

[0165] Fig.11 The microphone 1031 corresponds to Fig.10 The microphone 110, Fig.11 The communicator 1050 corresponds to Fig.10 The communicator 140 in Fig.11 The processor 1010 corresponds to Fig.10 The processor 120 in Fig.11 The memory 1090 corresponds to Fig.10 The memory 130 in the embodiment of the present invention is described below, and therefore their detailed description is omitted here.

[0166] In one embodiment of the present disclosure, the tuner 1040 may tune and select a frequency of a channel received only by the electronic device 1000 among various radio wave components by performing amplification, mixing, or resonance on a broadcast signal received via wire or wirelessly. The broadcast signal includes audio data, video, and additional information such as an electronic program guide (EPG).

[0167] The tuner 1040 may receive a broadcast signal from various sources, for example, terrestrial broadcasting, cable broadcasting, satellite broadcasting, Internet broadcasting, etc. The tuner 1040 may receive a broadcast signal from a source such as an analog broadcasting or a digital broadcasting.

[0168] In one embodiment of the present disclosure, the communicator 1050 may send data or signals to an external device or server and receive data or signals from an external device or server under the control of the processor 1010. The processor 1010 may send content to or receive content from an external device (not shown) connected thereto through the communicator 1050, download an application from the external device, or perform web browsing.

[0169] In addition, the communicator 1050 may include a combination including one or more of a wireless LAN module 1051 , a Bluetooth module 1052 , and a wired Ethernet module 1053 .

[0170] The video processor 1080 processes video data received by the electronic device 1000. The video processor 1080 may perform various image processing operations, such as decoding, scaling, noise filtering, frame rate conversion, resolution conversion, etc., on the video data.

[0171] The detector 1030 may detect a user's voice, a user's image, or a user's interaction, and includes a microphone 1031 , a camera 1032 , and an optical receiver 1033 .

[0172] The microphone 1031 receives a voice uttered by the user. The microphone 1031 may convert the received voice into an electrical signal and output the electrical signal to the processor 1010. The user's voice may include, for example, a voice corresponding to a menu or function of the electronic device 1000.

[0173] The camera 1032 may receive images (e.g., continuous frames) corresponding to the user's actions (including gestures) within the camera recognition range. The processor 1010 may select a menu displayed on the electronic device 1000 by using the received recognition result of the action, or may perform control corresponding to the recognition result of the action.

[0174] The optical receiver 1033 receives an optical signal (including a control signal) from an external control device through an optical window (not shown) of a frame of the display 1020, etc. The optical receiver 1033 may receive an optical signal corresponding to a user input (e.g., touch, press, touch gesture, voice, or motion) from a control device (not shown). Under the control of the processor 1010, a control signal may be extracted from the received optical signal.

[0175] Under the control of the processor 1010, the input / output 1070 receives a video (e.g., a moving image), an audio signal (e.g., voice or music), additional information (e.g., EPG), etc. from outside the electronic device 1000. Alternatively, under the control of the processor 1010, the input / output 1070 transmits a video, an audio signal, additional information, etc. to an external device. The input / output 1070 may include one of a high-definition multimedia interface (HDMI) port 1071, a component jack 1072, a PC port 1073, and a USB port 1074. The input / output 1070 may include a combination of the HDMI port 1071, the component jack 1072, the PC port 1073, and the USB port 1074. In addition, the input / output 1070 may include a display port (DP), a thunderbolt port, etc.

[0176] The processor 1010 controls the overall operation of the electronic device 1000 and the signal flow between the internal components of the electronic device 1000, and performs data processing functions. The processor 1010 can execute an operating system and various applications stored in the memory 1090 when there is a user input or a preset and stored condition is met.

[0177] The processor 1010 may include a random access memory (RAM) and a read-only memory (ROM), wherein the random access memory (RAM) is used to store signals or data input from the outside of the electronic device 1000, or is used as a storage area corresponding to various operations performed by the electronic device 1000, and the read-only memory (ROM) is used to store a control program for controlling the electronic device 1000.

[0178] The processor 1010 may include a graphics processor (not shown). The graphics processor generates a screen including various objects (e.g., icons, images, and texts) using an operation unit (not shown) and a renderer (not shown). The operation unit calculates attribute values ​​of each object to be displayed according to the layout of the screen, such as coordinates, shape, size, and color, by using the user input detected by the detector 1030. The renderer generates screens including various layouts of objects based on the attribute values ​​calculated by the operation unit. The screen generated by the renderer is displayed on the display area of ​​the display 1020.

[0179] The display 1020 generates a driving signal by converting an image signal, a data signal, an on-screen display (OSD) signal, a control signal, etc. processed by the processor 1010. The display 1020 may be implemented as a plasma display panel (PDP), a liquid crystal panel (LCD), an organic light emitting diode (OLED) display, a flexible display, etc., or may be implemented as a three-dimensional (3D) display. Alternatively, the display 1020 may be configured as a touch screen and used as an input device in addition to being used as an output device.

[0180] The audio processor 1085 processes audio data. The audio processor 1085 may perform various operations on the audio data, such as decoding, amplification, and noise filtering. The audio processor 1085 may include a plurality of audio processing modules to process audio data corresponding to a plurality of pieces of content.

[0181] Under the control of the processor 1010, the audio outputter 1060 outputs audio data included in the broadcast signal received through the tuner 1040. The audio outputter 1060 may output audio data (e.g., voice or sound) input via the communicator 1050 or the inputter / outputter 1070. Under the control of the processor 1010, the audio outputter 1060 may output audio data stored in the memory 1090. The audio outputter 1060 may include at least one of a speaker 1061, an earphone output terminal 1062, or a Sony / Philips Digital Interface (S / PDIF) output terminal 1063. The audio outputter 1060 may include a combination of the speaker 1061, the earphone output terminal 1062, and the S / PDIF output terminal 1063.

[0182] Under the control of the processor 1010, the power supply 1095 supplies power from an external power source to the internal components of the electronic device 1000. Alternatively, under the control of the processor 1010, the power supply 1095 supplies power output from one or more batteries (not shown) included in the electronic device 1000 to the internal components.

[0183] Under the control of the processor 1010, the memory 1090 may store various data, programs or applications for driving and controlling the electronic device 1000. Although not shown, the memory 1090 may include a broadcast receiving module, a channel control module, a volume control module, a communication control module, a voice recognition module, a motion recognition module, an optical receiving module, a display control module, an audio control module, an external input control module, a power control module, a power control module of an external device of a wireless connection (such as Bluetooth), a voice database (DB) or an action database (DB). The modules and databases (not shown) of the memory 1090 may be implemented as software to perform broadcast receiving control functions, channel control functions, volume control functions, communication control functions, voice recognition functions, motion recognition functions and optical receiving control functions, display control functions, audio control functions, external input control functions, power control functions or power control functions of external devices of a wireless connection (such as Bluetooth) by the electronic device 1000. The processor 1010 may perform each of these functions by using the software stored in the memory 1090.

[0184] Fig.10 and Fig.11The block diagrams of the electronic devices 100 and 1000 shown in are only block diagrams according to embodiments of the present disclosure. According to the description of the electronic devices 100 and 1000 actually implemented, some components of each block diagram may be combined together or omitted, or new components may be added thereto. That is, two or more components may be combined into one component, or one component may be divided into two or more components as needed. The functions performed in each block are provided only for explaining embodiments of the present disclosure, and the scope of the present disclosure should not be interpreted as being limited to a specific operation or device.

[0185] According to the embodiments of the present disclosure, the operation method of the electronic device can be implemented in the form of program instructions executed by various computer means and recorded on a computer-readable recording medium. The computer-readable recording medium may include separate program instructions, data files, data structures, etc. or a combination thereof. The program instructions recorded on the medium may be specially designed and configured for the present disclosure, or may be well known and available to ordinary technicians in the field of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as optical magnetic floppy disks, and hardware devices such as ROMs, RAMs and flash memories, which are specially configured to store and execute program instructions. Examples of program instructions include not only machine codes generated by compilers, but also high-level language codes executable by computers using interpreters, etc.

[0186] The operating method of the electronic device according to the embodiment of the present disclosure set forth herein may be provided by being included in a computer program product. The computer program product may be a product for transaction between a seller and a buyer.

[0187] A computer program product may include a software program and a computer-readable storage medium on which the software program is stored. For example, a computer program product may include a product (e.g., a downloadable application) that is electronically distributed in the form of a software program by a broadcast receiving device manufacturer or an electronic market (e.g., Google Play Store or App Store). For electronic distribution, at least a portion of the software program may be stored in a storage medium or temporarily generated. In this case, the storage medium may be a storage medium of a manufacturer's server, a server of an electronic market, or a relay server that temporarily stores the software program.

[0188] The computer program product may include a storage medium of a server or a storage medium of a client device in a system including a server and a client device. Alternatively, when there is a third device (e.g., a smartphone) communicatively connected to the server or the client device, the computer program product may include a storage medium of the third device. Alternatively, the computer program product may include a software program transmitted from the server to the client device or the third device or from the third device to the client device.

[0189] In this case, one of the server, the client device, and the third device may execute a computer program product to perform the method according to the embodiment of the present disclosure set forth herein. Alternatively, two or more of the server, the client device, and the third device may execute a computer program product to perform the method according to the embodiment of the present disclosure set forth herein in a distributed manner.

[0190] For example, a server (such as a cloud server or an artificial intelligence server) may execute a computer program product stored therein to control a client device communicatively connected to the server to perform the method according to an embodiment of the present disclosure set forth herein.

[0191] In one embodiment of the present disclosure, the electronic device may apply different parameters to perform voice activity detection according to the user's speech characteristics, thereby improving the accuracy and efficiency of voice activity detection.

[0192] It is possible to prevent speech loss of a user with a low speech rate and reduce unnecessary waiting time of a user with a high speech rate.

[0193] User utterance characteristics can be easily analyzed by using utterances of trigger words.

[0194] Although the embodiments of the present disclosure have been described in detail above, the scope of the present disclosure is not limited thereto, and it should be understood that various modifications and improvements made by ordinary technicians in the field using the basic concepts of the present disclosure defined in the appended claims are included in the scope of the present disclosure.

Claims

1. Electronic devices, including: a memory storing one or more instructions; as well as a processor configured to execute the one or more instructions stored in the memory to: Receiving audio data corresponding to a user utterance, wherein the audio data includes an utterance portion, the utterance portion including a trigger word indicating the start of the user utterance; Identifying whether the user speech corresponds to the trigger word; Analyzing the user speech corresponding to the trigger word to identify user speech features, the user speech features including at least one of speech rate, speech energy level, or pronunciation voice height; changing a parameter for performing a function corresponding to the voice detection by using the recognized user utterance feature; and By using the changed parameters, the function corresponding to the voice detection is performed on the user utterance input after the trigger word is received.

2. The electronic device according to claim 1, wherein: The processor is further configured to execute the one or more instructions stored in the memory to compare the recognized user speech feature with a reference speech feature regarding the trigger word, and change the parameter based on a result of the comparison.

3. The electronic device according to claim 1, wherein: The processor is also configured to execute the one or more instructions stored in the memory to: detect the starting point of the user's speech in the received audio data and the end point of the user's speech in the received audio data when the user speaks, so as to distinguish the speech part of the received audio data from the non-speech part of the received audio data.

4. The electronic device according to claim 1, wherein: The parameters for performing a function corresponding to voice detection include at least one of an energy threshold, a delay time, or an automatic end time for identifying the speech portion of the received audio data when the user speaks. 5 . The electronic device of claim 1 , further comprising a communicator configured to receive the audio data corresponding to the user utterance. 6 . The electronic device of claim 1 , further comprising a microphone configured to receive the user utterance and convert the user utterance into the audio data.

7. A method for operating an electronic device, the method comprising: Receiving audio data corresponding to a user utterance, wherein the audio data includes an utterance portion, the utterance portion including a trigger word indicating the start of the user utterance; Identifying whether the user speech corresponds to the trigger word; Analyzing the user speech corresponding to the trigger word to identify user speech features, the user speech features including at least one of speech rate, speech energy level, or pronunciation voice height; changing a parameter for performing a function corresponding to the voice detection by using the recognized user utterance feature; and By using the changed parameters, the function corresponding to the voice detection is performed on the user's utterance input after the trigger word is received.

8. The operating method according to claim 7, wherein: Changing the parameter for performing the function corresponding to the voice detection includes comparing the recognized user speech feature with a reference speech feature regarding the trigger word, and changing the parameter based on a result of the comparison.

9. The operating method according to claim 7, wherein: Performing the function corresponding to voice detection on the user utterance includes: detecting the starting point and the ending point of the user utterance when the user speaks, so as to distinguish the speech part of the received audio data from the non-speech part of the received audio data. 10 . A non-transitory computer-readable recording medium on which a program for executing the operating method according to claim 7 in a computer is recorded.

Images