Electronic device and Method for controlling the electronic device thereof

Abstract

An electronic device and a method for controlling the same are provided. The electronic device of the present disclosure can determine whether to activate a speech recognition model based on whether a wake-up word is included in a first sound signal input through a microphone and a second sound signal output through a speaker.

Description

Technology Field

[0001] The present disclosure relates to an electronic device and a method for controlling the same, and more specifically, to an electronic device and a method for controlling the same that activate a speech recognition model based on a sound signal output from the electronic device and a sound signal input through a microphone. Background Technology

[0002] Conventional speech recognition devices perform speech recognition processing using sound sources input through a microphone. However, the sound sources input into the microphone may include not only sound sources input through user speech but also sound sources output from the device's speaker. Therefore, conventionally, Acoustic Echo Cancellation (AEC) technology was utilized to distinguish between sound sources from user speech input into the microphone and sound sources output from the device's speaker.

[0003] AEC technology is a technique that converts the sound source input into the microphone and the sound source output from the speech recognition device into the frequency domain, and then extracts only the sound source of the voice produced by the user's speech by subtracting the signal value of the voice output from the speech recognition device from the signal value of the sound source input into the microphone.

[0004] However, AEC technology has a limitation in that the voice output from the voice recognition device and the voice input into the microphone must be synchronized in time. Additionally, since synchronizing each voice to utilize AEC technology increases the computational load, there is a limitation in that the development and manufacturing costs of voice recognition devices may increase. The problem to be solved

[0005] The present disclosure is devised to solve the aforementioned problems, and the purpose of the present disclosure is to provide an electronic device and a method for controlling the same that analyze a sound signal output from an electronic device and a sound signal input through a microphone to identify whether to activate a speech recognition model. means of solving the problem

[0006] According to one embodiment of the present disclosure for achieving the above-described purpose, an electronic device comprises a speaker, a microphone, a memory for storing a voice recognition model, and a processor. The processor inputs a second sound signal output through the speaker into a wake-up word detection model while a first sound signal is input through the microphone to identify whether the second sound signal contains a wake-up word, and if it is identified that the second sound signal does not contain the wake-up word, inputs the first sound signal input through the microphone into the wake-up word detection model to identify whether the first sound signal contains the wake-up word.

[0007] Meanwhile, a control method for an electronic device including a memory storing a voice recognition model according to an embodiment for achieving the above-described purpose may include the step of inputting a second sound signal output through a speaker into a wake-up word detection model while a first sound signal is input through a microphone to identify whether the second sound signal contains a wake-up word, and, if it is identified that the second sound signal does not contain the wake-up word, inputting the first sound signal input through the microphone into the wake-up word detection model to identify whether the first sound signal contains the wake-up word.

[0008] And, according to an embodiment for achieving the above-described purpose, an electronic device comprises a speaker, a microphone, a memory for storing at least one voice recognition model, and a processor, wherein the processor inputs a first sound signal input through the microphone to a first wake-up word detection model corresponding to each of the at least one voice recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of the at least one voice recognition model, and if it is identified that the first sound signal contains one of the wake-up words, inputs a second sound signal output through the speaker to a second wake-up word detection model corresponding to one of the wake-up words to identify whether the second sound signal contains one of the wake-up words.

[0009] In addition, a control method for an electronic device storing at least one voice recognition model according to an embodiment for achieving the above-described purpose may include the steps of: inputting a first sound signal input through a microphone to a first wake-up word detection model corresponding to each of the at least one voice recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of the at least one voice recognition model; and, if it is identified that the first sound signal contains one of the wake-up words, inputting a second sound signal output through a speaker to a second wake-up word detection model corresponding to one of the wake-up words to identify whether the second sound signal contains one of the wake-up words. Effects of the invention

[0010] According to various embodiments of the present disclosure described above, by distinguishing between a user voice signal input through a microphone and a wake-up word included in a voice signal output from the electronic device, the user can efficiently utilize voice recognition technology without utilizing AEC technology. Brief explanation of the drawing

[0011] FIGS. 1A and 1B are drawings for illustrating a method in which an electronic device activates a voice recognition model according to one embodiment of the present disclosure. FIG. 2a is a block diagram briefly illustrating the configuration of an electronic device according to one embodiment of the present disclosure. FIG. 2b is a block diagram illustrating in detail the configuration of an electronic device according to one embodiment of the present disclosure. FIG. 3 is a flowchart for explaining the operation method of an electronic device according to one embodiment of the present disclosure, FIG. 4 is a flowchart for explaining another mode of operation of an electronic device according to one embodiment of the present disclosure, FIG. 5 is a flowchart for explaining a method for controlling an electronic device according to one embodiment of the present disclosure, FIG. 6 is a flowchart for explaining another control method of an electronic device according to one embodiment of the present disclosure. Specific details for implementing the invention

[0012] Hereinafter, embodiments according to the present disclosure are described in detail with reference to the attached drawings so that those skilled in the art can easily implement them.

[0013] FIGS. 1a and 1b are drawings illustrating a method for an electronic device (100) to activate a voice recognition model according to one embodiment of the present disclosure. Meanwhile, as shown in FIGS. 1a and 1b, the electronic device (100) may be implemented as a display device, but this is merely one embodiment, and the electronic device (100) may include at least one of a tablet PC, a desktop PC, a laptop PC, a netbook computer, a server, a medical device, or a wearable device. Also, according to another embodiment of the present disclosure, the electronic device (100) may be implemented as at least one of a refrigerator, an air conditioner, an air purifier, a set-top box, or a media box (e.g., Samsung HomeSync™, Apple TV™, or Google TV™). And, as shown in FIGS. 1a and 1b, the microphone (120) may be implemented as a configuration included in a remote control (200) that controls the electronic device (100), but this is merely one embodiment, and it is obvious that it may be embedded in the electronic device (100).

[0014] The electronic device (100) can output a sound signal through a speaker (110), and the output sound signal can be input to a microphone (120) that is included in the remote control (200) or built into the electronic device (100). Meanwhile, the microphone (120) may receive a first sound signal that includes not only the sound signal output from the electronic device (100) but also various sound signals generated around the microphone (120) (e.g., the voice of a user located around the microphone (120)). As illustrated in FIGS. 1a and 1b, when the microphone (120) is included in the remote control (200), the first sound signal input to the microphone (120) can be input to a processor (140) within the electronic device (100) via a wired / wireless interface (e.g., Wi-Fi, Bluetooth module, etc.).

[0015] Meanwhile, in another embodiment of the present disclosure, the electronic device (100) may receive a first sound signal from a user terminal device that runs an application to control the electronic device (100). Specifically, the electronic device (100) may be connected to the user terminal device via various communication methods, such as Wi-Fi Direct or Bluetooth. Furthermore, when the first sound signal is received through a microphone included in the user terminal device while the application to control the electronic device (100) is running, the user terminal device may transmit the first sound signal to the electronic device (100). That is, the electronic device (100) may receive the first sound signal not only through a built-in microphone (120) or a microphone (120) included in a remote control (200), but also from a user terminal device that runs an application to control the electronic device (100). Meanwhile, the user terminal device may be implemented as various electronic devices such as a smartphone, a tablet PC, or a laptop.

[0016] As illustrated in FIG. 1a, the electronic device (100) can output a second sound signal (e.g., “Hi Bixby”) (10) containing a wake-up word through a speaker (110). Meanwhile, the electronic device (100) can store the second sound signal (10) in a storage configuration within the electronic device (100), such as a register or volatile memory within the processor (140). Then, the electronic device (100) can output the second sound signal (10) through the speaker (110) simultaneously with storing it or within a critical time. That is, the electronic device (100) can distinguish between the first sound signal and the second sound signal by storing the second sound signal (10) output through the speaker (110) in a register or volatile memory, etc., and storing the first sound signal input through the microphone (120) included in the remote control (200) in another area of ​​the register or volatile memory.

[0017] Meanwhile, the wake-up word is a word that can activate a speech recognition model, which is an artificial intelligence model capable of providing a response to a user's voice stored in the electronic device (100), and may also be expressed as a trigger word or a call word. The wake-up word may be a word that is predefined when the electronic device (100) is manufactured, and it is obvious that it can be added, changed, or deleted by the user. Furthermore, the activation of the speech recognition model may mean the action of loading the speech recognition model stored in non-volatile memory into volatile memory so that it can perform speech recognition functions.

[0018] Meanwhile, while the first sound signal is input through the microphone (120), the electronic device (100) can input the second sound signal (10) output through the speaker (110) into a wake-up word detection model to identify whether the second sound signal (10) contains a wake-up word. If the microphone (120) is included in the remote control (200), a delay may occur while the first sound signal input to the microphone (120) is input to the electronic device (100) through the communication module of the remote control (200). Accordingly, in one embodiment, the electronic device (100) can identify whether the second sound signal (10) contains a wake-up word while the delay occurs. In another embodiment, the electronic device (100) may perform an operation related to the first sound after identifying whether the second sound signal (10) contains a wake-up word, even if the first sound signal is input from a remote control (200) that includes a microphone (120). In another embodiment, the electronic device (100) may start an operation to identify whether the second sound signal (10) contains a wake-up word and perform an operation related to the first sound after a threshold time.

[0019] Meanwhile, the wake-up detection model stored in the electronic device (100) can identify whether the first and second sound signals contain a predefined wake-up word using keyword spotting technology, etc.

[0020] As illustrated in FIG. 1a, if it is identified that the second sound signal (10) contains a wake-up word, the electronic device (100) can be controlled not to activate the voice recognition model. Specifically, the wake-up word included in the sound signal output from the speaker (110) may not be a wake-up word spoken by the user for the purpose of activating the voice recognition model. Therefore, if it is identified that the second sound signal (10) output through the speaker (110) contains a wake-up word, the electronic device (100) can be controlled not to activate the voice recognition model even if the first sound signal contains a wake-up word.

[0021] In one embodiment, if it is identified that the second sound signal (10) does not contain a wake-up word, the electronic device (100) can input the first sound signal input through the microphone (120) into a wake-up word detection model to identify whether the first sound signal contains a wake-up word. In another embodiment, after starting the operation to identify whether the second sound signal (10) contains a wake-up word and after a threshold time, the electronic device (100) can input the first sound signal into a wake-up detection model to identify whether the first sound signal contains a wake-up word. At this time, if it is identified that the second sound signal (10) contains a wake-up word, the electronic device (100) can stop the operation to identify whether the first sound signal contains a wake-up word and control the voice recognition model so that it is not activated.

[0022] And, if it is identified that the first sound signal contains a wake-up word, the electronic device (100) can activate the voice recognition model. That is, the wake-up word contained in the first sound signal is not a wake-up word output through the speaker (110) of the electronic device (100), but may be a wake-up word spoken by a user for the purpose of activating the voice recognition model. Thus, as shown in FIG. 1b, if the second sound signal (10) does not contain a wake-up word but the first sound signal (20) contains a wake-up word, the electronic device (100) can activate the voice recognition model.

[0023] And, as illustrated in FIG. 1b, when the voice recognition model is activated, the electronic device (100) may display an indicator (30) indicating that the voice recognition model is activated or a message (40) asking the user for a request. However, this is merely one embodiment, and the electronic device (100) may display various indicators or UIs indicating that the voice recognition model is activated.

[0024] Meanwhile, in one embodiment, the electronic device (100) may control the parameters for performing operations included in the wake-up word detection model to input the first sound signal and the wake-up word detection model to input the second sound signal so that they are different from each other. In the case of the first sound signal input through the microphone (120), it may contain interference effects, reverberation effects, etc., due to a relatively large amount of noise compared to the second sound signal output through the speaker (120). Therefore, when identifying whether the first sound signal contains a wake-up word, the electronic device (100) may change the parameters of the wake-up word detection model so that the wake-up word can be identified while removing noise, reverberation effects, etc. However, this is merely one embodiment, and the detection model for detecting the wake-up word included in the first sound signal and the detection model for detecting the wake-up word included in the second sound signal may be implemented as separate models.

[0025] Additionally, the electronic device (100) may store one or more voice recognition models. If there is one or more voice recognition models, the wake-up words that can activate each of the one or more voice recognition models may also be different. For example, if the wake-up word that can activate the first voice recognition model is “Hi Bixby” as shown in FIG. 1a and FIG. 1b, the wake-up word that can activate the second voice recognition model may be “Hello Moogle”. Additionally, a wake-up detection model capable of identifying the wake-up word corresponding to one or more voice recognition models may be implemented as a different model. Thus, when multiple voice recognition models are stored, the electronic device (100) may input the second sound signal (10) into all wake-up word detection models corresponding to each of the multiple voice recognition models to identify whether the second sound signal contains one of the wake-up words corresponding to each of the multiple voice recognition models.

[0026] And, if it is identified that the second sound signal does not contain a wake-up word corresponding to each of the plurality of voice recognition models, the electronic device (100) inputs the first sound signal to all wake-up word detection models corresponding to each of the plurality of voice recognition models to identify whether the first sound signal contains one of the wake-up words corresponding to each of the plurality of voice recognition models. If it is identified that the first sound signal contains one of the wake-up words, the electronic device (100) can activate the voice recognition model corresponding to the wake-up word. On the other hand, if it is identified that the first sound signal does not contain a wake-up word, the electronic device (100) can control that one or more voice recognition models are not activated. That is, even when multiple voice recognition models are stored, the electronic device (100) can identify whether the wake-up word input through the microphone (120) is a wake-up word that matches the user's intention to activate the voice recognition model.

[0028] Meanwhile, in another embodiment of the present disclosure, the electronic device (100) can identify whether the first sound input through the microphone (120) contains a wake-up word before identifying whether the second sound signal output through the speaker (110) contains a wake-up word.

[0029] As described above, when one or more voice recognition models are stored, the electronic device (100) may store a wake-up word detection model capable of identifying a wake-up word capable of activating each of the one or more voice recognition models. Accordingly, the electronic device (100) may input a first sound signal (10) input through a microphone (120) into a first wake-up word detection model corresponding to each of at least one voice recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of the one or more voice recognition models. That is, the electronic device (100) may first identify whether the first sound signal (10) contains one of the wake-up words corresponding to the one or more voice recognition models, while storing the second sound signal in a storage space such as a register or volatile memory.

[0030] When it is identified that one of the wake-up words is included in the first sound signal, the electronic device (100) can input the second sound signal (10) output through the speaker (110) into a second wake-up word detection model corresponding to one of the wake-up words identified as being included in the first sound signal, thereby identifying whether the second sound signal includes one of the corresponding wake-up words. And, when it is identified that one of the wake-up words is included in the second sound signal, the electronic device (100) can control the voice recognition model so that it is not activated. That is, the electronic device (100) can input the second sound signal only into one second wake-up word detection model corresponding to a specific wake-up word identified as being included in the first sound signal, rather than inputting it into all of the second wake-up word detection models corresponding to one or more voice recognition models. And, when it is identified that a specific wake-up word is included in the second wake-up word detection model, since a sound signal containing the specific wake-up word is input through the speaker (110), the electronic device (100) can be controlled not to activate the voice recognition model.

[0031] For example, if it is identified that the first sound signal contains one of the wake-up words, ‘Hi Bixby’, the electronic device (100) can input the second sound signal to a second wake-up word detection model capable of identifying and detecting ‘Hi Bixby’ among at least one second wake-up word detection model. Then, if it is identified through the detection model that the second sound signal does not contain ‘Hi Bixby’, the electronic device (100) can activate a voice recognition model.

[0032] Meanwhile, if it is identified that the first sound signal does not contain one of the wake-up words corresponding to at least one voice recognition model, the electronic device (100) may be controlled not to activate the voice recognition model. That is, since the first sound signal input through the microphone (120) does not contain a wake-up word, it may mean that the user did not utter the wake-up word, so the electronic device (100) may not activate the voice recognition model.

[0034] FIG. 2a is a block diagram briefly illustrating the configuration of an electronic device (100) according to one embodiment of the present disclosure. As shown in FIG. 2a, the electronic device (100) may include a speaker (110), a microphone (120), a memory (130), and a processor (140). However, the configuration shown in FIG. 2a is an illustrative diagram for implementing embodiments of the present disclosure, and appropriate hardware and software configurations that are obvious to a person skilled in the art may additionally be included in the electronic device (100).

[0035] The speaker (110) is configured to output various audio data, as well as various notification sounds or voice messages, after various processing operations such as decoding, amplification, and noise filtering have been performed by the audio processing unit. Meanwhile, the configuration for outputting audio can be implemented as a speaker, but this is merely one embodiment, and it can be implemented as an output terminal capable of outputting audio data. Also, the second sound signal to be output from the speaker (110) may be stored in a register or volatile memory, etc., under the control of the processor (140).

[0036] The microphone (120) is configured to receive a first sound signal containing an ambient sound signal of the electronic device (100). In one embodiment, the microphone (120) may be included in a remote control (200) capable of controlling the electronic device (100). In this case, the first sound signal input to the microphone (120) may be input to a processor (140) via a wired / wireless interface (e.g., Wi-Fi, Bluetooth module, etc.). In another embodiment, the microphone (120) may be built into the electronic device (100).

[0037] The memory (130) may store instructions or data related to at least one other component of the electronic device (100). In particular, the memory (130) may be implemented as non-volatile memory (e.g., Dynamic Random-Access Memory (RAM) and Static RAM), volatile memory (Flash Memory, Programmable Read-Only Memory (PROM), Magnetoresistive Random-Access Memory (MRAM) and Resistive RAM (RRAM)), a hard disk drive (HDD), or a solid-state drive (SSD).

[0038] Meanwhile, volatile memory refers to memory that requires a continuous power supply to maintain stored information. Volatile memory may be implemented as a separate component from the processor (140) that can be linked with the processor (140), but this is merely one embodiment, and volatile memory may be implemented as a component of the processor (140) and included in the processor (140). Furthermore, non-volatile memory refers to memory that can maintain stored information even if the power supply is interrupted.

[0039] The memory (130) is accessed by the processor (140), and the processor (140) may perform reading, writing, modifying, deleting, updating, etc. of data. In the present disclosure, the term memory may include the memory (130), ROM (not shown), RAM (not shown) within the processor (140), or a memory card (not shown) (e.g., micro SD card, memory stick) mounted in the electronic device (100). Additionally, the memory (130) may store programs and data, etc., for configuring various screens to be displayed in the display area of ​​the display.

[0040] In particular, the memory (130) can store a speech recognition model, which is an artificial intelligence model capable of providing a response to a user's voice. The speech recognition model is stored in non-volatile memory and can be loaded into volatile memory and activated by the control of the processor (140) when a wake-up word spoken by the user is identified. The speech recognition model outputs text corresponding to the user's voice and can determine the intent of the output text.

[0041] Also, the memory (130) can store a wake-up word detection model. The wake-up word detection model stored in the non-volatile memory can be loaded into the volatile memory and activated by the control of the processor (140) when the power of the electronic device (100) is turned on. Thus, when the power of the electronic device (100) is turned on, the wake-up word detection model is activated and can identify the wake-up word contained in the first sound signal and the second sound signal.

[0042] Meanwhile, a plurality of voice recognition models may be stored in the memory (130). In one embodiment, a wake-up word detection model corresponding to each of the plurality of voice recognition models may be stored in the memory (130). In another example, a first wake-up word detection model and a second wake-up word detection model corresponding to each of the plurality of voice recognition models may be stored in the memory (130). The first wake-up word detection model is a model that detects and identifies a wake-up word included in a first sound signal input through the microphone (120), and the second wake-up word detection model is a model that detects and identifies a wake-up word included in a second sound signal output through the speaker (110).

[0043] The processor (140) is electrically connected to the memory (130) and can control the overall operation and function of the electronic device (100). In particular, while a first sound signal is input through the microphone (120), the processor (140) can input a second sound signal output through the speaker (110) into a wake-up word detection model to identify whether the second sound signal contains a wake-up word.

[0044] If it is identified that the second sound signal contains a wake-up word, the processor (140) can control the voice recognition model not to be activated even if the first sound signal contains a wake-up word. That is, since the wake-up word contained in the first sound signal output through the speaker (110) is not a wake-up word spoken by the user to activate the voice recognition model, the processor (140) can control the voice recognition model not to be activated.

[0045] Meanwhile, if it is identified that the second sound signal does not contain a wake-up word, the processor (140) inputs the first sound signal input through the microphone (120) into a wake-up detection model to identify whether the first sound signal contains a wake-up word. In one embodiment, if it is identified that the first sound signal contains a wake-up word, the processor (140) can activate a speech recognition model. Meanwhile, in one embodiment, if it is identified that the first sound signal does not contain a wake-up word, the processor (140) can control the speech recognition model so that it is not activated.

[0046] And, when a plurality of voice recognition models are stored in memory (130), the processor (140) inputs a second sound signal to all wake-up word detection models corresponding to each of the plurality of voice recognition models to identify whether one of the wake-up words corresponding to each of the plurality of voice recognition models is included in the second sound signal.

[0047] If it is identified that the second sound signal does not contain a wake-up word corresponding to each of the plurality of speech recognition models, the processor (140) inputs the first sound signal to a wake-up word detection model corresponding to each of the plurality of speech recognition models to identify whether the first sound signal contains one of the wake-up words corresponding to each of the plurality of speech recognition models. If it is identified that the first sound signal contains one of the wake-up words corresponding to each of the plurality of speech recognition models, the processor (140) can activate the speech recognition model. As another example, if it is identified that the first sound signal does not contain a wake-up word, the processor (140) can control the speech recognition model so that it is not activated.

[0048] Meanwhile, if it is identified that the second sound signal contains one of the wake-up words corresponding to each of the plurality of speech recognition models, the processor (140) can control the voice recognition model so that it is not activated even if the first sound signal contains one of the wake-up words corresponding to the plurality of speech recognition models.

[0050] Meanwhile, in another embodiment of the present disclosure, the processor (140) inputs a first sound signal input through the microphone (120) to a first wake-up word detection model corresponding to each of at least one speech recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of at least one speech recognition model.

[0051] If it is identified that the first sound signal does not contain one of the wake-up words corresponding to at least one speech recognition model, the processor (140) can control the speech recognition model not to be activated. That is, since the first sound signal input through the microphone (120) does not contain a wake-up word, it may mean that the user did not utter a wake-up word to activate the speech recognition model, so the processor (140) can control the speech recognition model not to be activated.

[0052] Meanwhile, if it is identified that one of the wake-up words is included in the first sound signal, the processor (140) can input the second sound signal output through the speaker (110) into a second wake-up word detection model corresponding to one of the identified wake-up words to identify whether one of the identified wake-up words is included in the second sound signal. That is, since the case where the first sound signal input through the microphone (120) contains a wake-up word may include the case where the second sound signal output through the speaker (110) contains a wake-up word, the processor (140) can identify whether one of the identified wake-up words is also included in the second sound signal. For example, if it is identified that one of the identified wake-up words is included in the second sound signal through the second wake-up word detection model corresponding to one of the wake-up words, the processor (140) can control the voice recognition model not to be activated. As another example, if it is identified that one of the wake-up words is not included in the second sound signal, the processor (140) can control the voice recognition model to be activated.

[0054] Additionally, the processor (140) may include or be defined by one or more of a central processing unit (CPU) that processes digital signals, a Micro Controller Unit (MCU), a micro processing unit (MPU), a controller, an application processor (AP), a communication processor (CP), or an ARM processor. Furthermore, the processor (140) may be implemented as a System on Chip (SoC) or Large Scale Integration (LSI) with built-in processing algorithms, or may be implemented in the form of a Field Programmable Gate Array (FPGA). The processor (140) can perform various functions by executing computer executable instructions stored in memory (130). Moreover, to perform artificial intelligence functions, the processor (140) may include at least one of a separate AI-dedicated processor such as a graphics-processing unit (GPU), a Neural Processing Unit (NPU), or a Visual Processing Unit (VPU).

[0056] FIG. 2b is a block diagram illustrating in detail the configuration of an electronic device (100) according to one embodiment of the present disclosure. As shown in FIG. 2b, the electronic device (100) may include a speaker (110), a microphone (120), a memory (130), a processor (140), a communication unit (150), a display (160), an input unit (170), a sensor (180), and an input / output interface (190). Meanwhile, since the speaker (110), microphone (120), memory (130), and processor (140) have been described in detail with reference to FIG. 2a, a redundant description will be omitted.

[0057] The communication unit (150) includes a circuit and can perform communication with an external device. At this time, the communication unit (150) may communicate with the external device by passing through a third device (e.g., a medium device, a hub, an access point, a server, or a gateway). In particular, the communication unit (150) can receive a second sound signal from the remote control (200) through various communication modules, such as a Bluetooth module, and input and transmit it to the processor (140).

[0058] Meanwhile, the communication unit (150) may include various communication modules to perform communication with an external device. For example, the communication unit (150) may include a wireless communication module, such as LTE, LTE-A (LTE Advance), CDMA (code division multiple access), WCDMA (wideband CDMA), 5G (5 th It may include a cellular communication module using at least one of Generation, UMTS (universal mobile telecommunications system), WiBro (Wireless Broadband), or GSM (Global System for Mobile Communications). As another example, the wireless communication module may include at least one of, for example, WiFi (wireless fidelity), Bluetooth, Bluetooth Low Energy (BLE), Zigbee, NFC (near field communication), Magnetic Secure Transmission, Radio Frequency (RF), or Body Area Network (BAN).

[0059] The display (160) can display various information under the control of the processor (140). In particular, when the voice recognition model is activated, the display (160) can display an indicator that the voice recognition model has been activated under the control of the processor (140). Additionally, the display (160) can display a message or UI requesting a user voice to be input into the voice recognition model under the control of the processor (140).

[0060] Additionally, the display (160) can be implemented as a touch screen with a touch panel. Furthermore, the panel of the display (160) can be implemented with various display technologies such as LCD (Liquid Crystal Display), OLED (Organic Light Emitting Diodes), AM-OLED (Active-Matrix Organic Light-Emitting Diode), LcoS (Liquid Crystal on Silicon), QLED (Quantum dot Light-Emitting Diode), and DLP (Digital Light Processing).

[0061] The input unit (170) can receive user input for controlling the electronic device (100). In particular, the input unit (170) may include a touch panel for receiving user touch input using a user's hand or a stylus pen, a button for receiving user operation input, etc. Additionally, the input unit (170) may be implemented as another input device (e.g., a keyboard, a mouse, a motion input unit, etc.).

[0062] The sensor (180) may include a sensor capable of detecting environmental information around the electronic device (100) (e.g., a temperature sensor, a humidity sensor, a barometric pressure sensor, etc.), a sensor capable of detecting the presence of a user (e.g., a camera, a UWB sensor, an IR sensor, a proximity sensor, a light sensor, etc.). In addition, the sensor (180) may further include an image sensor for capturing the outside of the electronic device (100).

[0063] The input / output interface (190) can input or output at least one of audio and video signals. In particular, the input / output interface (190) can receive video including at least one of content and UI from an external device and can output control commands to the external device. Meanwhile, the input / output interface (190) may be HDMI (High Definition Multimedia Interface), but this is merely one embodiment, and it may be any one of MHL (Mobile High-Definition Link), USB (Universal Serial Bus), DP (Display Port), Thunderbolt, VGA (Video Graphics Array) port, RGB port, D-SUB (D-subminiature), and DVI (Digital Visual Interface). Depending on the implementation example, the input / output interface (190) may include a port for inputting and outputting only audio signals and a port for inputting and outputting only video signals as separate ports, or it may be implemented as a single port for inputting and outputting both audio and video signals. Meanwhile, although the communication unit (150) and the input / output interface (190) are shown as separate components, this is merely one embodiment, and the input / output interface (190) can also be implemented in a form included in the communication unit (150).

[0065] FIG. 3 is a first flowchart for explaining the operation of an electronic device (100) according to an embodiment of the present disclosure. First, the electronic device (100) can output a second sound signal through a speaker (110) (S310). The electronic device (100) outputs the second sound signal through the speaker (110) after performing operations such as decoding, amplification, and noise filtering by an audio processing unit, and the second sound signal to be output can be stored in a storage space such as a register or memory. Then, while the first sound signal is input through the microphone (110), the electronic device (100) can input the second sound signal into a wake-up detection model to identify whether the second sound signal contains a wake-up word (S320).

[0066] If it is identified that the second sound signal contains a wake-up word, the electronic device (100) can control the voice recognition model so that it is not activated (S320-Y). Meanwhile, if it is identified that the second sound signal does not contain a wake-up word, the electronic device (100) can input the first sound signal into the wake-up word detection model to identify whether the first sound signal contains a wake-up word (S320-N). Meanwhile, the electronic device (100) can control the parameters of the wake-up word detection model to input the first sound signal to be changed. That is, the electronic device (100) can change the computational parameters of the wake-up word detection model to input the first sound signal to remove noise filtering and reverberation effects.

[0067] Meanwhile, if it is identified that the first sound signal does not contain a wake-up word, the electronic device (100) can control the voice recognition model so that it is not activated (S320-Y). Then, if it is identified that the first sound signal contains a wake-up word, the electronic device (100) can activate the voice recognition model (S330).

[0068] In the above embodiment, during the delay time when the first sound signal input through the microphone (120) is transmitted to the electronic device (100) through a wired / wireless interface, the electronic device (100) identifies whether the second sound signal contains a wake-up word, and since it identifies whether the first sound signal contains a wake-up word, it can determine whether to activate the voice recognition model without wasting time compared to the embodiment according to the second flowchart described later.

[0070] FIG. 4 is a second flowchart for explaining the operation of an electronic device (100) according to one embodiment of the present disclosure.

[0071] First, the electronic device (100) can store the second sound signal to be output through the speaker (110) in memory or a register, etc. (S410). Then, the electronic device (100) can input the first sound signal input through the microphone (120) into a first wake-up word detection model corresponding to each of at least one voice recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of at least one voice recognition model (S420).

[0072] If the first sound signal does not contain a wake-up word corresponding to at least one speech recognition model, the electronic device (100) can control the speech recognition model so that it is not activated (S420-N). Meanwhile, if it is identified that the first sound signal contains one of the wake-up words corresponding to at least one speech recognition model, the electronic device (100) can identify whether the second sound signal contains one of the identified wake-up words (S420-Y). Specifically, the electronic device (100) can identify whether the second sound signal contains one of the identified wake-up words by inputting the second sound signal to a detection model corresponding to one of the identified wake-up words among one or more second wake-up word detection models.

[0073] If it is determined that one of the identified wake-up words is included in the second sound signal, the electronic device (100) can control the voice recognition model so that it is not activated (S420-N). And, if it is determined that one of the identified wake-up words is not included in the second sound signal, the electronic device (100) can activate the voice recognition model (S430).

[0074] In the above embodiment, when a plurality of voice recognition models are stored, the electronic device (100) first identifies which wake-up word is included in the first sound signal input through the microphone (120), and inputs the second sound signal only to the detection model corresponding to the identified wake-up word among the plurality of second wake-up word detection models to determine the user's intention. That is, the electronic device (100) can use fewer processor (140) and memory (130) resources compared to the embodiment according to the first flowchart described above.

[0076] FIG. 5 is a first flowchart for explaining a method of controlling an electronic device (100) according to one embodiment of the present disclosure.

[0077] First, while a first sound signal is input through the microphone (120), the electronic device (100) inputs a second sound signal output through the speaker (110) into a wake-up word detection model to identify whether the second sound signal contains a wake-up word (S510). The electronic device (100) stores the second sound signal to be output through the speaker (110), and since the first sound signal is input through a wired / wireless interface, the first sound signal and the second sound signal can be distinguished.

[0078] And, if it is identified that the second sound signal does not contain a wake-up word, the electronic device (100) can input the first sound signal input through the microphone (120) into a wake-up word detection model to identify whether the first sound signal contains a wake-up word (S520). If the first sound signal contains a wake-up word, the electronic device (100) can activate the voice recognition model, but if the first sound signal does not contain a wake-up word, the electronic device (100) can control the voice recognition model so that it is not activated.

[0080] FIG. 6 is a second flowchart for explaining a method of controlling an electronic device (100) according to one embodiment of the present disclosure.

[0081] The electronic device (100) inputs a first sound signal input through a microphone to a first wake-up word detection model corresponding to each of at least one voice recognition model to identify whether the first sound signal contains one of the wake-up words corresponding to each of at least one voice recognition model (S610). Then, if it is identified that the first sound signal contains one of the wake-up words, the electronic device (100) inputs a second sound signal output through a speaker (110) to a second wake-up word detection model corresponding to one of the wake-up words to identify whether the second sound signal contains one of the wake-up words. If it is identified that the second sound signal contains one of the identified wake-up words, the electronic device (100) can control the voice recognition model so that it is not activated. Then, if it is identified that the second sound signal does not contain one of the identified wake-up words, the electronic device (100) can control the voice recognition model to be activated.

[0083] Meanwhile, the drawings attached to this disclosure are not intended to limit the technology described in this disclosure to specific embodiments and should be understood to include various modifications, equivalents, and / or alternatives to the embodiments of this disclosure. In connection with the description of the drawings, similar reference numerals may be used for similar components.

[0084] In the present disclosure, expressions such as “have,” “may have,” “include,” or “may include” indicate the presence of such features (e.g., numerical values, functions, actions, or components such as parts) and do not exclude the presence of additional features.

[0085] In the present disclosure, expressions such as “A or B,” “at least one of A or / and B,” or “one or more of A or / and B” may include all possible combinations of items listed together. For example, “A or B,” “at least one of A and B,” or “at least one of A or B” may refer to cases including (1) at least one A, (2) at least one B, or (3) both at least one A and at least one B.

[0086] Expressions such as "first," "second," "first," or "second" used in this disclosure may modify various components regardless of order and / or importance, and are used only to distinguish one component from another and do not limit said components.

[0087] When it is stated that a component (e.g., a first component) is "(operatively or communicatively) coupled with" or "connected to" another component (e.g., a second component), it should be understood that the component may be directly connected to the other component or connected through another component (e.g., a third component). On the other hand, when it is stated that a component (e.g., a first component) is "directly connected" or "directly connected" to another component (e.g., a second component), it may be understood that no other component (e.g., a third component) exists between the component and the other component.

[0088] As used in this disclosure, the expression “configured to” may be replaced, depending on the context, with, for example, “suitable for,” “having the capacity to,” “designed to,” “adapted to,” “made to,” or “capable of.” The term “configured to” may not necessarily mean only “specifically designed to” in hardware. Instead, in some contexts, the expression “device configured to” may mean that the device is “capable of” in conjunction with other devices or components. For example, the phrase “subprocessor configured to perform A, B, and C” may mean a dedicated processor for performing the said operations (e.g., an embedded processor), or a generic-purpose processor (e.g., a CPU or an application processor) capable of performing said operations by executing one or more software programs stored in a memory device.

[0089] Meanwhile, the term "user" may refer to a person using an electronic device or a device using an electronic device (e.g., an artificial intelligence electronic device). The present disclosure will be described in more detail below with reference to the drawings.

[0090] Various embodiments of the present disclosure may be implemented as software containing instructions stored on a machine-readable storage medium (e.g., a computer). The machine may include an electronic device (e.g., an electronic device (100)) according to the disclosed embodiments, which is a device capable of calling instructions stored from the storage medium and operating according to the called instructions. When the instructions are executed by a processor, the processor may perform a function corresponding to the instructions directly or using other components under the control of the processor. The instructions may include code generated or executed by a compiler or an interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, "non-transitory storage medium" means that it does not contain a signal and is tangible, without distinguishing whether data is stored semi-permanently or temporarily on the storage medium. For example, the "non-transitory storage medium" may include a buffer in which data is temporarily stored.

[0091] According to one embodiment, the method according to the various embodiments disclosed herein may be provided as included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed online in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the computer program product (e.g., a downloadable app) may be temporarily stored or temporarily created in a storage medium such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0092] Each component (e.g., module or program) according to various embodiments may be composed of a singular or multiple entities, and some of the aforementioned sub-components may be omitted, or other sub-components may be additionally included in various embodiments. Generally or additionally, some components (e.g., module or program) may be integrated into a single entity to perform the functions performed by each of the respective components prior to integration in the same or similar manner. The operations performed by the module, program, or other components according to various embodiments may be executed sequentially, in parallel, iteratively, or heuristically, or at least some operations may be executed in a different order, omitted, or other operations added. Explanation of the symbols

[0093] 110: Speaker 120: Microphone 130: Memory 140: Processor

Claims

Claim 1 An electronic device comprising: a microphone for receiving a first sound signal; a speaker for outputting a second sound signal; a memory for storing a voice recognition model; and a processor; wherein the processor controls the voice recognition function corresponding to the voice recognition model to be activated when the first sound signal contains a wake-up word and the second sound signal does not contain the wake-up word, and controls the voice recognition function corresponding to the voice recognition model not to be activated when the second sound signal contains the wake-up word. Claim 2 delete Claim 3 delete Claim 4 An electronic device according to claim 1, wherein, when a plurality of voice recognition models are stored in the memory, the processor inputs the second sound signal to a wake-up word detection model corresponding to each of the plurality of voice recognition models to identify whether the second sound signal contains one of the wake-up words corresponding to each of the plurality of voice recognition models. Claim 5 An electronic device according to claim 4, wherein if the processor identifies that the second sound signal does not contain a wake-up word corresponding to each of the plurality of speech recognition models, the processor inputs the first sound signal to a wake-up word detection model corresponding to each of the plurality of speech recognition models to identify whether the first sound signal contains one of the wake-up words corresponding to each of the plurality of speech recognition models. Claim 6 An electronic device comprising: a speaker; a microphone; a memory storing a plurality of wake-up word detection models and a plurality of voice recognition models; wherein the plurality of wake-up word detection models and the plurality of voice recognition models correspond to each of the plurality of wake-up words, and a processor; wherein the processor inputs a first sound signal input through the microphone to each of the plurality of wake-up word detection models to identify whether the first wake-up word among the plurality of wake-up words is included in the first sound signal, and when the first wake-up word is identified as being included in the first sound signal, inputs a second sound signal to a first wake-up word detection model corresponding to the first wake-up word among the plurality of wake-up word detection models to identify whether the first wake-up word is included in a second sound signal output through the speaker. Claim 7 In claim 6, the processor is an electronic device that controls the plurality of speech recognition models not to be activated when it is identified through the first wake-up word detection model that the first wake-up word is included in the second sound signal. Claim 8 In claim 7, the processor is an electronic device that controls the plurality of voice recognition models not to activate them when it is identified that the first sound signal does not contain one of the wake-up words corresponding to the plurality of voice recognition models. Claim 9 A control method for an electronic device that receives a first sound signal through a microphone, outputs a second sound signal through a speaker, and stores a voice recognition model, comprising: a step of activating a voice recognition function corresponding to the voice recognition model when the first sound signal includes a wake-up word and the second sound signal does not include the wake-up word; and a step of controlling not to activate a voice recognition function corresponding to the voice recognition model when the second sound signal includes the wake-up word. Claim 10 delete Claim 11 delete Claim 12 In claim 9, the control method comprises the step of, when a plurality of voice recognition models are stored in the electronic device, inputting the second sound signal into a wake-up word detection model corresponding to each of the plurality of voice recognition models to identify whether one of the wake-up words corresponding to each of the plurality of voice recognition models is included in the second sound signal. Claim 13 In claim 12, the control method comprises the step of, when it is identified that the second sound signal does not contain a wake-up word corresponding to each of the plurality of speech recognition models, inputting the first sound signal to a wake-up word detection model corresponding to each of the plurality of speech recognition models to identify whether the first sound signal contains one of the wake-up words corresponding to each of the plurality of speech recognition models. Claim 14 A control method for an electronic device storing a plurality of wake-up word detection models and a plurality of voice recognition models, wherein the plurality of wake-up word detection models and the plurality of voice recognition models correspond to each of the plurality of wake-up words, and the control method inputs a first sound signal input through a microphone to each of the plurality of wake-up word detection models to identify whether the first wake-up word among the plurality of wake-up words is included in the first sound signal, and if the first wake-up word is identified as being included in the first sound signal, inputs a second sound signal to a first wake-up word detection model corresponding to the first wake-up word among the plurality of wake-up word detection models to identify whether the first wake-up word is included in a second sound signal output through a speaker. Claim 15 In claim 14, the control method further comprises the step of controlling the plurality of speech recognition models not to be activated when it is identified through the first wake-up word detection model that the first wake-up word is included in the second sound signal. Claim 16 In claim 14, the control method comprises the step of controlling not to activate the plurality of speech recognition models when it is identified that the first sound signal does not contain one of the wake-up words corresponding to the plurality of speech recognition models.

Images