KR-102962209-B1 - AN ARTIFICIAL INTELLIGENCE APPARATUS USING SOUND SIGNAL CLASSIFICATION AND METHOD FOR THE SAME

KR102962209B1KR 102962209 B1KR102962209 B1KR 102962209B1KR-102962209-B1

Abstract

A method for controlling an artificial intelligence device installed in a home appliance is proposed, wherein a processor controls an input unit to receive a sound signal at a preset time, removes noise from the received signal and separates it into a signal from the user and a signal from the device, and inputs the separated signals into an artificial intelligence model using multi-class classification to obtain a result value output by the artificial intelligence model. If the result value indicates user sleep, the sleep mode of the home appliance is executed and a sleep mode transition notification is output; if the result value indicates an unclear signal, feedback is requested from the user, and the artificial intelligence model is updated using the feedback.

Inventors

김재홍
정한길

Assignees

엘지전자 주식회사

Dates

Publication Date: 20260508
Application Date: 20190919

Claims (11)

In an artificial intelligence device installed in a home appliance, An input unit that receives a sound signal using a microphone; It includes a processor that controls the home appliance to use a sound signal received from the input unit as an input value for an artificial intelligence model, and executes a sleep mode when the result of the artificial intelligence model is determined to be user sleep. The above artificial intelligence model includes a multi-class classification model that classifies the user's non-sleep, user's sleep, device signals, and unclear signals as result values based on previously learned data. The above multi-class classification model includes a neural network with adjusted weights, which is trained by using sound signals received from the input unit as input values and labeling user sleep, user non-sleep, and unclear signals as output values. Artificial intelligence device.
delete
delete
In Article 1, The processor removes noise from the sound signal received from the input unit, separates the noise-removed signal into a signal by the user and a device signal, and uses the separated signals as input values for the artificial intelligence model. Artificial intelligence device.
In paragraph 1 The above processor uses the sound signal received by the input unit as the input value of the artificial intelligence model during the sleep detection mode execution time set by the user. Artificial intelligence device.
In paragraph 5 The processor disables the sleep mode of the home appliance when the result of the artificial intelligence model based on the sound signal received from the input unit after the user's preset wake-up time has passed is user sleep. Artificial intelligence device.
In paragraph 1 If the result of the artificial intelligence model is an unclear signal, the processor requests feedback from the user regarding whether the user was asleep or awake during that time, and updates the artificial intelligence model using the feedback data. Artificial intelligence device.
In paragraph 1 It further includes a communication unit that communicates with at least one home appliance. The above processor When the result of the artificial intelligence model is user sleep, the communication unit controls at least one home appliance to execute sleep mode. Artificial intelligence device.
In paragraph 1 An artificial intelligence device further comprising an output unit that displays a sleep mode switching notification of the home appliance when the result of the artificial intelligence model is user sleep.
A method for controlling an artificial intelligence device installed in a home appliance, Step of receiving a sound signal at a preset time; A step of removing noise from the received signal and separating it into a signal by the user and a signal by the device; A step of inputting the separated signal into an artificial intelligence model using multi-class classification to obtain a result value output by the artificial intelligence model; If the result value is user sleep, the method includes the step of executing the sleep mode of the home appliance and outputting a sleep mode switching notification, and if the result value is an unclear signal, requesting feedback from the user and updating the artificial intelligence model using the feedback. The above artificial intelligence model includes a multi-class classification model that classifies the user's non-sleep, user's sleep, device signals, and unclear signals as result values based on previously learned data. The above multi-class classification model includes a neural network with adjusted weights, which is trained by using a sound signal received from an input unit as an input value and labeling user sleep, user non-sleep, and unclear signals as output values.
In a recording medium having a program recorded thereon for performing a method of controlling an artificial intelligence device installed in a home appliance, The method for controlling the above artificial intelligence device Step of receiving a sound signal at a preset time; A step of removing noise from the received signal and separating it into a signal by the user and a signal by the device; A step of inputting the separated signal into an artificial intelligence model using multi-class classification to obtain a result value output by the artificial intelligence model; If the result value is user sleep, the method includes the step of executing the sleep mode of the home appliance and outputting a sleep mode switching notification, and if the result value is an unclear signal, requesting feedback from the user and updating the artificial intelligence model using the feedback. The above artificial intelligence model includes a multi-class classification model that classifies the user's non-sleep, user's sleep, device signals, and unclear signals as result values based on previously learned data. A recording medium comprising a neural network with adjusted weights, wherein the above multi-class classification model is trained by using sound signals received from an input unit as input values and labeling user sleep, user non-sleep, and unclear signals as output values.

Description

An Artificial Intelligence Apparatus Using Sound Signal Classification and Method for the Same The present invention relates to an artificial intelligence device and method for operating a sleep mode of a home appliance based on the classification of received sound signals. Specifically, the present invention relates to an artificial intelligence device and method that classifies sound signals received in daily life by various household devices, such as IoT devices like air conditioners, air purifiers, boilers, refrigerators, lighting fixtures, motorized blinds, robot vacuum cleaners, and AI speakers, determines whether a user is in a sleep state, and switches to a sleep mode if the user is in a sleep state. Recently, research and applications of artificial intelligence using deep learning algorithms have been active. Artificial intelligence models based on deep learning algorithms are trained using training data. Each training dataset includes labeled data, which serves as the output value of the artificial intelligence model. Among these algorithms, multi-class classification models are used as algorithms for classification that distinguish three or more classes. In relation to the present invention, there have been smart devices that use wearable devices or mobile devices attached to a user's body to infer a sleep state through changes in the user's activity level, ambient light, and heart rate, and evaluate sleep time and sleep quality based on collected sensor information. However, these devices only evaluate the user's sleep state and have the problem of not being able to improve the user's sleep environment by actually controlling surrounding devices. FIG. 1 shows an artificial intelligence device (100) according to one embodiment of the present invention. FIG. 2 shows an artificial intelligence server (200) according to one embodiment of the present invention. FIG. 3 shows an artificial intelligence system (1) according to one embodiment of the present invention. FIG. 4 is an overall flowchart according to one embodiment of the present invention. FIG. 5 is an artificial intelligence model according to one embodiment of the present invention. FIG. 6 is an artificial intelligence model according to one embodiment of the present invention. FIG. 7 is a flowchart of a personalized learning sequence according to one embodiment of the present invention. FIG. 8 is a general mode switching flowchart according to one embodiment of the present invention. Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components regardless of drawing symbols will be assigned the same reference number, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the present invention. Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between. Artificial Intelligence (AI) Artificial intelligence refers to the field of researching artificial intelligence or the methodologies to create it, while machine learning refers to the field of researching methodologies to define and solve various problems addressed within the field of artificial intelligence. Machine learning is also defined as an algorithm that improves performance on a task through continuous experience. An Artificial Neural Network (ANN) is a model used in machine learning that can refer to any model capable of problem-solving, composed of artificial neurons (nodes) that form a network through the connection of synapses. A