Noisy recognition processing method and device

Abstract

The invention discloses a noisy recognition processing method and device. The method comprises the following steps: acquiring a sound signal in at least one room in a house to obtain a current sound signal, and identifying the current sound signal to determine whether there is a noisy sound; under the condition that the existence of the noisy sound is determined, sound changing processing is carried out on the noisy sound to obtain a sound changing transmission sound, the sound changing processing represents reduction of the speed of the noisy sound, and the timbre of the noisy sound is kept unchanged; and a plurality of central control devices are adopted to rebroadcast the variable sound transmission sound, and one central control device is installed in each room in the house. According to the invention, the technical problem of low propagation efficiency caused by the fact that information transmission among different rooms is noisy by adopting manual transmission or traditional communication equipment in the existing scheme is solved.

Description

Technical Field

[0001] The present invention relates to the field of smart home technology, and in particular to a quarrel recognition and processing method and device. Background Art

[0002] With the continuous development of smart home technology, the central control screen, as the core control terminal of home smart devices, has increasingly rich functions and application scenarios. However, in family life, due to various reasons, family members may quarrel over certain things, which not only affects family harmony, but also may have adverse effects on the physical and mental health of family members. Therefore, it is necessary to develop an intelligent central control screen system that can monitor family quarrels and intervene in time.

[0003] In traditional home environments, information transmission between different rooms often relies on manual transmission or traditional communication equipment, which is inefficient and prone to missing important information. Traditional smart home systems lack the ability to intervene in family quarrels and can only passively record or transmit information.

[0004] That is to say, the existing scheme uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, which makes the transmission efficiency low, and may aggravate the tense atmosphere at the quarrel scene.

[0005] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0006] The embodiment of the present invention provides a method and device for processing quarrel identification, so as to at least solve the technical problem that the existing solution uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, thereby resulting in low transmission efficiency.

[0007] According to one aspect of an embodiment of the present invention, a method for identifying and processing a quarrel is provided, comprising: acquiring a sound signal in at least one room in a house to obtain a current sound signal, and identifying the current sound signal to determine whether there is a quarrel sound; if it is determined that there is a quarrel sound, performing voice-changing processing on the quarrel sound to obtain a voice-changed propagation sound, wherein the voice-changing processing is characterized by reducing the speech speed of the quarrel sound while keeping the timbre of the quarrel sound unchanged; and using multiple central control devices to broadcast the voice-changed propagation sound, wherein one central control device is respectively installed in each of the rooms in the house.

[0008] Optionally, the current sound signal is identified to determine whether there is a quarrel sound, including: preprocessing the current sound signal to obtain a target sound signal, and performing feature extraction on the target sound signal to obtain feature data, wherein the preprocessing includes deduplication processing and denoising processing; using a machine learning model to identify the feature data to determine whether there is the quarrel sound, wherein the machine learning model includes a support vector machine model, a random forest model and a deep learning model.

[0009] Optionally, after determining that there is a quarrel sound, the quarrel identification and processing method further includes: using a machine learning model to process the feature data to determine a quarrel sound pattern, wherein the quarrel sound pattern represents the intensity of the quarrel; using the multiple central control devices to play preset music or preset sentences that conform to the quarrel sound pattern, wherein the preset music is music with a rhythm within a preset range, and the preset sentence is a sentence with a speaking speed lower than a preset speaking speed.

[0010] Optionally, after acquiring the sound signal in at least one room in the house to obtain the current sound signal, the quarrel identification and processing method further includes: using speech recognition technology to convert the current sound signal into text data; preprocessing the text data to obtain target text data, wherein the preprocessing includes deduplication processing and denoising processing; using natural language processing technology, at least based on the syntax of the target text data, the relationship between different speakers, and the emotions of each speaker, processing the target text data to determine the cause of the quarrel.

[0011] Optionally, after determining the cause of the quarrel, the quarrel identification and processing method further includes: based on the cause of the quarrel, combining the natural language processing technology and the machine learning model to process the current sound signal, and generating and using the multiple central control devices to play family activity suggestions.

[0012] Optionally, the quarrel identification and processing method further includes: optimizing parameters of the natural language processing technology and parameters of the machine learning model using a K-fold validation method.

[0013] Optionally, after using multiple central control devices to relay the voice-changing propagation sound, the quarrel identification and processing method further includes: controlling the multiple central control devices to display the time, location and intensity of the quarrel in text form respectively.

[0014] According to another aspect of an embodiment of the present invention, a device for identifying a family quarrel is provided, comprising: an acquisition unit, configured to acquire a sound signal in at least one room in a house, obtain a current sound signal, and identify the current sound signal to determine whether there is a quarrel sound; a first processing unit, configured to, when it is determined that there is a quarrel sound, perform voice-changing processing on the quarrel sound to obtain a voice-changed propagation sound, wherein the voice-changing processing is characterized by reducing the speech speed of the quarrel sound while keeping the timbre of the quarrel sound unchanged; and a second processing unit, configured to broadcast the voice-changed propagation sound using a plurality of central control devices, wherein one central control device is installed in each of the rooms in the house.

[0015] Optionally, the acquisition unit includes: a preprocessing subunit, used to preprocess the current sound signal to obtain a target sound signal, and perform feature extraction on the target sound signal to obtain feature data, wherein the preprocessing includes deduplication processing and denoising processing; an identification subunit, used to use a machine learning model to identify the feature data to determine whether the quarrel sound exists, and the machine learning model includes a support vector machine model, a random forest model and a deep learning model.

[0016] Optionally, the device for identifying family quarrels further includes: a third processing unit, used to use a machine learning model to process the feature data after determining the presence of quarrel sounds, so as to determine a quarrel sound pattern, wherein the quarrel sound pattern represents the intensity of the quarrel; a playing unit, used to use the multiple central control devices to play preset music or preset sentences that conform to the quarrel sound pattern, wherein the preset music is music with a rhythm within a preset range, and the preset sentence is a sentence with a speaking speed lower than a preset speaking speed.

[0017] Optionally, the device for identifying family quarrels also includes: a conversion unit, used to obtain sound signals in at least one room in the house, and then use speech recognition technology to convert the current sound signal into text data after obtaining the current sound signal; a preprocessing unit, used to preprocess the text data to obtain target text data, wherein the preprocessing includes deduplication processing and denoising processing; a fourth processing unit, used to use natural language processing technology to process the target text data based on at least the syntax of the target text data, the relationship between different speakers, and the emotions of each speaker, so as to determine the cause of the quarrel.

[0018] Optionally, the device for identifying family quarrels further includes: a fifth processing unit, for combining the natural language processing technology and the machine learning model to process the current sound signal based on the cause of the quarrel after determining the cause of the quarrel, and generating and using the multiple central control devices to play family activity suggestions.

[0019] Optionally, the device for identifying family quarrels further includes: an optimization unit, configured to optimize the parameters of the natural language processing technology and the parameters of the machine learning model by using a K-fold validation method.

[0020] Optionally, the device for identifying family quarrels further includes: a control unit, which is used to control the multiple central control devices to display the time, place and intensity of the quarrel in text form after the multiple central control devices are used to broadcast the voice-changing sound.

[0021] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned dispute identification and processing methods.

[0022] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of the above-mentioned quarrel identification and processing methods.

[0023] According to another aspect of an embodiment of the present invention, a processor is further provided, and the processor is used to run a program, wherein the program executes any one of the above-mentioned dispute identification and processing methods when running.

[0024] According to another aspect of an embodiment of the present invention, a computer program product is provided, comprising computer instructions, and when the computer instructions are executed by a processor, any one of the above-mentioned dispute identification and processing methods is executed.

[0025] In an embodiment of the present invention, a sound signal in at least one room in a house is obtained to obtain a current sound signal, and the current sound signal is identified to determine whether there is a quarrel sound; if it is determined that there is a quarrel sound, the quarrel sound is processed by voice change to obtain a voice change propagation sound, and the voice change processing is characterized by reducing the speech speed of the quarrel sound and keeping the timbre of the quarrel sound unchanged; multiple central control devices are used to broadcast the voice change propagation sound, wherein a central control device is installed in each room in the house. Through the technical solution provided by the present invention, the screen-to-screen linkage technology is adopted, and a sound recognition module is built into the central control screen to accurately identify the quarrel sound, and the notification module is used to remind other family members of the occurrence of the quarrel. At the same time, the voice change processing module is used to perform voice change processing to reduce the intensity of the quarrel, thereby achieving the purpose of information exchange between family members, real-time monitoring and identification of family quarrels, and improving the efficiency and accuracy of the transmission, thereby solving the technical problem that the existing solution uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, thereby making the transmission efficiency low. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0027] Figure 1 It is a hardware structure block diagram of a mobile terminal of a method for processing quarrel identification according to an embodiment of the present invention;

[0028] Figure 2 is a flowchart of a quarrel identification and processing method according to an embodiment of the present invention;

[0029] Figure 3 is a flow chart of an optional quarrel identification and processing method according to an embodiment of the present invention;

[0030] Figure 4 is a schematic diagram of a device for identifying a family quarrel according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] As introduced in the background technology, the existing solution uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, which makes the transmission efficiency low. In view of the above defects, in the embodiments of the present invention, a quarrel identification and processing method and device, electronic device, computer readable storage medium, processor and computer program product are provided.

[0034] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0035] The method embodiments provided in the embodiments of the present invention can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for processing a quarrel identification according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0036] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the quarrel identification and processing method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above network include but are not limited to the Internet, corporate intranets, local area networks, mobile communication networks and combinations thereof. The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0037] According to an embodiment of the present invention, a method embodiment of a quarrel identification and processing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] Figure 2 is a flow chart of a method for processing quarrel identification according to an embodiment of the present invention. Figure 2 As shown, the method comprises the following steps:

[0039] Step S202, obtaining a sound signal in at least one room in the house, obtaining a current sound signal, and identifying the current sound signal to determine whether there is a quarrel sound.

[0040] In this embodiment, obtaining the sound signal in at least one room in the house can help the system monitor the sound conditions in the room in real time and obtain the current sound signal to facilitate subsequent real-time analysis and processing. By identifying the current sound signal, the system can determine whether there is a quarrel sound, so as to take corresponding measures in time.

[0041] Figure 3 is a flowchart of an optional quarrel identification and processing method according to an embodiment of the present invention. Figure 3 As shown, it mainly includes three systems, namely, voice recognition and quarrel monitoring system, screen linkage and reminder system, and voice change processing and harmony reminder system. Specifically, it includes multiple central control screens distributed in different rooms of the home, voice recognition modules, linkage control modules, voice change processing modules and notification modules. Among them, the voice recognition and quarrel monitoring system collects sound signals in the room in real time through the microphone on the central control screen, and transmits them to the voice recognition module for processing. The voice recognition module analyzes and processes the collected sound signals, and determines whether there is a quarrel sound through the pre-trained quarrel recognition algorithm.

[0042] It should be noted that the above algorithm can be implemented based on audio features and machine learning models. When the sound recognition module recognizes the sound of quarreling, the quarrel monitoring mechanism will be triggered. At this time, the system will continuously monitor the quarreling sound.

[0043] Optionally, the above audio features may include: volume, pitch, speaking speed and other features.

[0044] Step S204, when it is determined that there is a quarrel sound, the quarrel sound is subjected to voice changing processing to obtain a voice-changed propagation sound, wherein the voice changing processing is characterized in that the speech speed of the quarrel sound is reduced while the timbre of the quarrel sound is kept unchanged.

[0045] In this embodiment, when it is determined that there is a quarrel sound, voice changing processing of the quarrel sound can help the recognition system to more accurately identify and analyze the quarrel sound, thereby improving the recognition accuracy of the quarrel sound. By reducing the speaking speed and keeping the timbre unchanged, the quarrel sound can be easier to distinguish and identify, while reducing the interference of other noises on the recognition results.

[0046] like Figure 3 As shown, while sending the quarrel information, the voice change processing module will change the quarrel sound to make it sound smoother or softer to reduce the intensity of the quarrel. Specifically, this can be achieved by adjusting parameters such as the frequency, amplitude or phase of the audio signal. The timbre is retained, but the tone can be slowed down, making the entire quarrel less intense, thereby slowing down the quarrel.

[0047] Step S206, using multiple central control devices to relay the voice-changing sound, wherein a central control device is installed in each room in the house.

[0048] In this embodiment, multiple central control devices are used to broadcast voice-changing sounds in each room, which can help the system to more accurately identify and analyze quarrels. By installing central control devices in each room, the system can monitor multiple sound sources at the same time, thereby improving the accuracy of quarrel recognition. By broadcasting sounds through voice changing, the system can better simulate sound conditions in different scenarios, thereby better identifying the occurrence of quarrels.

[0049] It should be noted that the above-mentioned central control device can be a central control screen, which is used to receive user voice commands or display information and is the core control terminal of home smart devices.

[0050] like Figure 3 As shown, when the sound recognition module recognizes the sound of quarreling, the linkage control module will trigger the linkage control and send the quarreling information to the central control screen in other rooms.

[0051] As can be seen from the above, in an embodiment of the present invention, a sound signal in at least one room in a house is obtained to obtain a current sound signal, and the current sound signal is identified to determine whether there is a quarrel sound; when it is determined that there is a quarrel sound, the quarrel sound is subjected to voice changing processing to obtain a voice-changed propagation sound, and the voice changing processing is characterized by reducing the speech speed of the quarrel sound and keeping the timbre of the quarrel sound unchanged; multiple central control devices are used to broadcast the voice-changed propagation sound, wherein a central control device is installed in each room in the house, so as to achieve the screen-to-screen linkage technology, and a sound recognition module is built into the central control screen to accurately identify the quarrel sound, and the notification module is used to remind other family members of the occurrence of the quarrel, and at the same time, the voice changing processing module is used to perform voice changing processing to reduce the intensity of the quarrel, thereby realizing information exchange between family members, real-time monitoring and identification of family quarrels, and improving the efficiency and accuracy of communication.

[0052] The technical solution provided by the embodiment of the present invention solves the technical problem that the existing solution uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, resulting in low transmission efficiency.

[0053] According to the above embodiment of the present invention, the current sound signal is identified to determine whether there is a quarrel sound, including: preprocessing the current sound signal to obtain a target sound signal, and extracting features of the target sound signal to obtain feature data, the preprocessing includes deduplication processing and denoising processing; using a machine learning model to identify the feature data to determine whether there is a quarrel sound, the machine learning model includes a support vector machine model, a random forest model and a deep learning model.

[0054] In this embodiment, the input sound signal is preprocessed to remove repeated data and noise to obtain a clear target sound signal, and features are extracted from the target sound signal to obtain feature data, which can be used to train a machine learning model. Machine learning models such as a support vector machine model, a random forest model, and a deep learning model are used to identify the feature data to determine whether there is a quarrel sound, and the quarrel sound can be effectively identified.

[0055] Specifically, audio collection is first performed. The system needs to collect the sounds in the home environment in real time through hardware devices such as microphone arrays, and use the sound recognition module built into each central control screen to monitor the sounds in the home environment in real time and identify quarreling sounds. The collected sounds are subjected to sound preprocessing operations such as denoising and enhancement to improve the accuracy of subsequent analysis. Feature extraction is then performed to extract key features from the preprocessed sounds, such as volume, pitch changes, speaking speed, and emotional characteristics (such as anger, sadness), etc. These features can be extracted through algorithms such as fast Fourier transform (FFT) and Mel-frequency cepstral coefficients (MFCC).

[0056] According to the above embodiment of the present invention, after determining that there is a quarrel sound, the quarrel identification and processing method further includes: using a machine learning model to process feature data to determine a quarrel sound pattern, the quarrel sound pattern characterizing the intensity of the quarrel; using multiple central control devices to play preset music or preset sentences that conform to the quarrel sound pattern, the preset music being music with a rhythm within a preset range, and the preset sentence being a sentence with a speaking speed lower than a preset speaking speed.

[0057] In this embodiment, processing feature data using a machine learning model can help determine the sound pattern of a quarrel, thereby characterizing the intensity of the quarrel. By analyzing and classifying the sound data, the occurrence and extent of the quarrel can be more accurately identified. Playing preset music or preset sentences that match the quarrel sound pattern through multiple central control devices can help reduce the emotion and intensity of the quarrel. The selection of preset music and sentences can affect people's emotions and behaviors. By playing music and sentences with a steady rhythm or a slower speed, the atmosphere of the quarrel can be eased and the possibility of conflict can be reduced.

[0058] Specifically, after feature extraction, quarrel detection processing is performed and the intensity of the quarrel is evaluated. The extracted sound features are analyzed using a machine learning model to identify the sound pattern of the quarrel. The model needs to be trained with a large amount of labeled quarrel sound data to improve the accuracy and robustness of the detection. The intensity of the quarrel can be divided into mild, moderate and severe levels according to some of the above-mentioned detection results. When the central control screen in a room detects the quarrel sound and reaches a preset threshold, a level between mild, moderate and severe can be set as the threshold, which will trigger the screen linkage mechanism. At this time, the central control screen will send quarrel information to the central server, including the time, place and intensity of the quarrel. After receiving the quarrel information, the central server will distribute it to the central control screens in other rooms. During the distribution process, the format and presentation of the information can be adjusted as needed to facilitate users to better understand and receive it.

[0059] In addition, the central control screen in other rooms can detect environmental information. If the environment is dark and relatively quiet, the screen can be bright and a relatively soothing and gentle prompt can be given. If the environment is relatively noisy, a louder prompt can be given.

[0060] It should be noted that the above-mentioned sound pattern is a comprehensive concept that covers a variety of sound features. These features together constitute the sound performance of a quarrel and are mainly used to judge the intensity of the quarrel in multiple dimensions. For example: 1) Volume feature. When quarreling, people's volume is often higher than that of normal communication, and may be accompanied by a sharp change in volume. Therefore, the volume and its rate of change are important features for identifying the sound pattern of the quarrel; 2) Tone feature. During the quarrel, due to emotional excitement, people's tone may become higher or lower, and may fluctuate frequently. The change pattern of tone is also one of the key features for identifying the sound pattern of the quarrel; 3) Speaking speed feature. When quarreling, people's speaking speed tends to speed up to express more opinions and emotions. Therefore, the speed of speech is also an important feature for judging 4) Sound quality features: During an argument, people’s voices may tremble, become hoarse, or have other quality changes due to emotional excitement and tension. These sound quality features also help identify the sound patterns of the argument. 5) Sound rhythm features: Sounds in an argument often have specific rhythmic patterns, such as frequent interruptions and overlapping speech. These rhythmic features are also important clues for identifying the sound patterns of the argument. 6) Emotional features: Although emotions themselves are not direct sound features, they can be indirectly inferred through certain elements of the sound (such as volume, pitch, speaking speed, etc.). Machine learning models can capture the relationship between these sound features and emotions by learning from a large amount of labeled quarrel sound data, thereby more accurately identifying the sound patterns of the argument.

[0061] In addition, it should be noted that the specific steps for selecting and training the above-mentioned machine learning model are as follows: 1) Model selection: Select a suitable machine learning model based on the specific application scenario and data characteristics. For quarrel detection, you can consider using a classification model or a sequence model; 2) Data annotation: Collect and annotate a large amount of family quarrel sound and text data, including quarrel samples of different intensities, to train the machine learning model; 3) Model training and optimization: Use the annotated data to train the model, and improve the performance of the model through methods such as cross-validation and parameter tuning.

[0062] The following uses LSTM as an example to illustrate the quarrel detection process. The specific steps are as follows: 1) Data preparation, including audio collection, data labeling, and feature extraction, that is, collecting audio data in a home environment through a microphone and other devices, and labeling the collected audio data as "quarrel" or "non-quarrel". Usually, it is necessary to manually monitor the audio and make judgments. Existing data sets can also be used for training to convert the audio data into a feature sequence suitable for LSTM processing, which usually includes segmenting the audio signal into frames and extracting Mel frequency cepstral coefficients (MFCC), short-time energy, zero-crossing rate and other features for each frame; 2) Data preprocessing, including standardization / normalization and sequence division, that is, standardizing or normalizing the extracted features to obtain the features. Ensuring that different features have the same scale is conducive to model training. The continuous audio feature sequence is divided into subsequences of fixed length. Each subsequence is used as an input sample of the LSTM model. If the length of the audio sequence is inconsistent, the length can be unified by truncation or padding; 3) Model construction, including LSTM layer, fully connected layer and activation function. The LSTM layer is built as the core part of the model to capture the temporal dependencies and long-term patterns in the audio sequence. The number of units in the LSTM layer (i.e., the size of the hidden layer) needs to be adjusted according to the specific task and data set. A fully connected layer (also called a dense layer or DNN layer) is added after the LSTM layer to map the output of the LSTM layer to the final classification result. The number of output nodes of the fully connected layer should match the number of categories (i.e., "quarrel" and "non-quarrel"). A softmax activation function is used at the output of the fully connected layer to output the predicted probability of each category. 4) Model training, including loss function, optimizer, and training process. The cross entropy loss function is selected as the loss function of the model because it is suitable for multi-classification problems. Optimizers such as Adam and RMSprop are used to update the weights of the model to minimize the loss function. The labeled audio feature sequence is input into the model, the loss is calculated through forward propagation, and then the back propagation algorithm is used to update the weights of the model. Multiple iterations (i.e., multiple epochs) are required until the performance of the model reaches the optimal value. No longer significantly improved or reaches the preset training rounds; 5) Model evaluation, which may include validation sets and performance indicators. Using a portion of the data set that did not participate in the training (validation set) to evaluate the performance of the model helps to avoid overfitting and understand the performance of the model on unseen data. The model's accuracy, recall rate, F1 score and other performance indicators on the validation set are calculated to evaluate the classification effect of the model; 6) Deployment and application, which may include real-time detection and intervention strategies. The trained LSTM model is deployed to the smart central control screen system, which receives audio data in the home environment in real time and detects quarrels. When the system detects a quarrel, it intervenes according to the preset intervention strategy to ease the tension.

[0063] Through the above steps, the LSTM model can be used to detect quarrels in the family environment, and corresponding intervention strategies can be formulated accordingly to promote harmonious relationships among family members.

[0064] Optionally, in the task of quarrel detection, you can choose to use sequence models in deep learning, especially long short-term memory networks (LSTM) or gated recurrent units (GRU), because they are good at processing time series data and capturing long-term dependencies in sequences, which is particularly important for analyzing continuous audio signals; the above-mentioned presentation methods may include but are not limited to: sound prompts, screen prompts, etc.; the above-mentioned classification models may include: SVM, RF, and deep learning classifiers, etc.; the above-mentioned sequence models may include: LSTM, GRU, etc., for processing time series data; the above-mentioned preset intervention strategies may include: playing soothing music, sending reminder messages, etc.

[0065] According to the above embodiment of the present invention, after acquiring the sound signal in at least one room in the house to obtain the current sound signal, the quarrel recognition and processing method further includes: using speech recognition technology to convert the current sound signal into text data; preprocessing the text data to obtain target text data, the preprocessing including deduplication processing and denoising processing; using natural language processing technology, at least based on the syntax of the target text data, the relationship between different speakers and the emotions of each speaker, processing the target text data to determine the cause of the quarrel.

[0066] In this embodiment, speech recognition technology is used to convert the quarrel content captured by voice signals into text data, and the text data is preprocessed to remove repetitions and noise to improve data quality. Natural language processing technology is then used to analyze the text data to identify syntax, relationships between different speakers, and emotions, thereby determining the cause of the quarrel and providing important information and support for subsequent conflict resolution.

[0067] It should be noted that for the voice content that can be captured, the natural language processing (NLP) model can be used for further semantic analysis to more accurately judge the intensity of the quarrel. The voice recognition function converts the collected sound into text, and uses the sentiment analysis technology in NLP to judge the emotional tendency of the converted text and identify expressions of negative emotions such as anger and dissatisfaction. After that, the text content is further analyzed to understand the specific reasons for the quarrel and the views of both parties, so as to carry out more precise intervention.

[0068] The specific steps are as follows: 1) Speech recognition and text conversion. First, the system needs to use speech recognition technology to convert the collected speech data into text data. The accuracy of this step directly affects the effect of subsequent semantic understanding. Therefore, the system needs to adopt a high-performance speech recognition engine to ensure that the speech is accurately converted into analyzable text; 2) Text preprocessing. The converted text data often contains some useless information, such as noise, repeated words, and irrelevant filler words. In order to improve the efficiency and accuracy of semantic understanding, the system needs to preprocess the text, including removing stop words, punctuation marks, numbers and other non-critical information, as well as word form restoration, part-of-speech tagging and other processing; 3) Semantic analysis. Based on the preprocessing, the system uses the semantic analysis technology in natural language processing (NLP) to conduct in-depth analysis of the text; 4) Opinion extraction and summary. On the basis of semantic analysis, the system needs to further extract opinion information from the text and summarize and generalize these opinions, which includes identifying the main opinions, arguments and rebuttals of both parties, and understanding the logical relationship between these opinions. Through this process, the system can grasp the overall picture of the quarrel more comprehensively and provide strong support for subsequent intervention; 5) Precise intervention strategy formulation. Based on the results of semantic understanding, the system can formulate more precise intervention strategies. For example, if the system identifies that the main cause of the quarrel is poor communication or misunderstanding, it can recommend that both parties communicate calmly or provide advice on communication skills. If the system finds that both parties are very emotional, it can choose to play soothing music or send soothing messages to calm emotions. Through precise intervention strategies, the system can effectively ease the tense atmosphere of family quarrels and promote harmonious relationships among family members.

[0069] Among them, semantic analysis mainly includes the following aspects: 1) Named entity recognition, identifying named entities in the text, such as names of people, places, and institutions, which are often related to the cause of the quarrel or the identities of both parties; 2) Syntactic analysis, parsing the syntactic structure of the text, understanding the relationship between the subject, predicate, object and other components in the sentence, so as to grasp the overall meaning of the sentence; 3) Sentiment analysis, judging the emotional tendency of the text, identifying the expression of negative emotions such as anger, dissatisfaction, and sadness, which are often closely related to the intensity and cause of the quarrel; 4) Topic detection and tracking, extracting key topics from the text and tracking the development trajectory of the topic in order to understand the specific content of the quarrel and the views of both parties.

[0070] According to the above embodiment of the present invention, after determining the cause of the quarrel, the quarrel identification and processing method further includes: based on the cause of the quarrel, combining natural language processing technology and machine learning models to process the current sound signal, generate and use multiple central control devices to play family activity suggestions.

[0071] In this embodiment, natural language processing technology and machine learning models are combined to process the current sound signal, which can help the system accurately analyze and understand the cause of the quarrel. By analyzing the voice content and emotional characteristics in the sound signal, the system can determine the specific cause of the quarrel. Generating and using multiple central control devices to play family activity suggestions can help ease the quarrel and provide some suggested activities or music to improve the family atmosphere. In this way, quarrels can be effectively resolved, harmonious relationships between family members can be promoted, and it can play a role in mediating quarrels and improving the family atmosphere.

[0072] According to the above embodiment of the present invention, the quarrel identification and processing method further includes: using a K-fold validation method to optimize parameters of the natural language processing technology and parameters of the machine learning model.

[0073] In this embodiment, the K-fold validation method can be used to optimize the parameters of the natural language processing technology and the parameters of the machine learning model. By dividing the data set into K parts, taking one of them as the validation set and the remaining K-1 parts as the training set, training the model multiple times and calculating the performance indicators on the validation set, a more reliable model performance evaluation result can be obtained. The K-fold validation method can help determine the optimal model parameters and improve the generalization ability of the model, thereby more accurately identifying and classifying controversial texts. It can also avoid overfitting and underfitting problems and improve the stability and reliability of the model.

[0074] It should be noted that the above-mentioned cross-validation is a technique for evaluating the performance of machine learning models. It divides the dataset into multiple small parts, and then trains and tests the model multiple times on these small parts to obtain a more comprehensive and reliable estimate of the model performance. In the training process of the quarrel detection model, cross-validation is particularly important because it can help us avoid overfitting and find the optimal model parameters.

[0075] Taking the use of cross-validation to train a quarrel detection model as an example, the embodiment is described, and the specific steps are as follows: 1) Dataset preparation: Assume that there is a data set containing a large amount of labeled audio data, which has been divided into two categories: "quarrel" and "non-quarrel". First, the data needs to be divided into a training set and a test set. The usual ratio is 70% as a training set and 30% as a test set. In this example, the application of cross-validation on the training set is mainly focused on; 2) Cross-validation division: For the training set, the cross-validation technology can be further used to divide it into multiple mutually exclusive subsets (folds). Common cross-validation methods include K-fold cross-validation (K-Fo ld Cross-Va l idat ion), where K is a hyperparameter representing the number of subsets. For example, if K=5 is selected, the training set will be divided into 5 subsets of equal size (if the dataset size is not divisible by K, the size of the subsets may need to be slightly adjusted); 3) Cross-validation process. In K-fold cross-validation, the model will be trained and validated K times, each time one of the subsets is selected as the validation set, and the remaining K-1 subsets are used as training sets; 4) Performance evaluation. After completing K rounds of training and validation, the average performance index of the model on the validation set in each round can be calculated to evaluate the overall performance of the model. This average provides a more robust estimate of how the model performs on different data subsets; 5) Parameter tuning. Cross-validation is not only used to evaluate the performance of the model, but also for parameter tuning. During the training process, different model parameters can be tried, and cross-validation can be used to compare the model performance under these parameter combinations. Finally, the parameter combination with the best performance is selected as the final model parameter.

[0076] Among them, the following are the specific steps of the cross-validation process: Round 1: select the first subset as the validation set, use the remaining four subsets to train the model, and then evaluate the performance of the model on the first subset; Round 2: select the second subset as the validation set, use the remaining subsets (including the subset used as the validation set in the first round) to train the model, and evaluate the performance on the second subset; Round K: select the last subset as the validation set, use all the previous subsets to train the model, and evaluate the performance on this subset.

[0077] Optionally, the above performance indicators may include: accuracy, recall rate, and F1 score, etc.; the above model parameters may include: the number of units in the LSTM layer, the learning rate, and the batch size, etc.

[0078] According to the above embodiment of the present invention, after using multiple central control devices to broadcast the voice-changing sound, the quarrel recognition and processing method further includes: controlling the multiple central control devices to display the time, location and intensity of the quarrel in text form.

[0079] In this embodiment, by controlling multiple central control devices to display the time, location and intensity of the quarrel, the occurrence of the quarrel can be understood in real time, a quick response can be made, key information of the quarrel can be presented more intuitively, and corresponding measures can be taken to deal with it.

[0080] Specifically, the central control screen that receives the quarrel information will remind other family members through text, voice or image. At the same time, the system can also send reminder information to mobile phones or other smart devices according to user settings. For example, you can set on the APP from the beginning that all quarrel information needs to be sent to the central controls under the same family. After the setting is completed, a central control can distribute the quarrel information to the pre-set central control after detecting it. The sound after voice changing can be sent to other family members together with other reminder information. In addition to direct reminder notifications, the system can also promote family harmony in other ways, for example, it can play some relaxing and pleasant music or videos to relieve the tense atmosphere; or provide some family activity suggestions to encourage family members to participate together and enhance their feelings.

[0081] Optionally, the reminder content may include: a brief description of the quarrel, suggested solutions or soothing words, etc.

[0082] like Figure 3 As shown, after receiving the quarrel information sent by the linkage control module, the notification module will notify other family members through the central control screen in other rooms by text, voice or image, reminding them to pay attention to family harmony.

[0083] It should be noted that the above reminders can be obtained through the following steps: 1) Case library construction, collecting a large number of family quarrel resolution cases and soothing words. These cases can come from the practical experience of psychological counseling experts, family therapists, social workers, etc., or from public family relationship forums, social media and other channels; 2) Data cleaning and annotation, cleaning the collected data, removing duplicate, irrelevant or low-quality content, and annotating each case, including the cause of the quarrel, the intensity, the effectiveness of the solution, and the applicability of the soothing words. 3) User feedback collection, the system should provide user feedback channels to collect family members' evaluations and suggestions on the generated solutions and soothing words; 4) Continuous optimization, based on user feedback and actual effect evaluation, continuously optimize the model algorithm and generation strategy to improve the accuracy and applicability of solutions and soothing words.

[0084] The technical solution provided by the above embodiment of the present invention solves the following problems: 1) In a traditional family environment, information transmission between rooms often relies on manual transmission or traditional communication equipment, which is inefficient and easy to miss important information; 2) The traditional smart home system lacks the ability to intervene in family quarrels and can only passively record or transmit information. At the same time, it has the following beneficial effects: 1) Through the screen-to-screen linkage technology, information exchange between different rooms in the family is realized, so that family members can understand the situation in the family in time, and the efficiency and accuracy of information transmission are improved; 2) After identifying the quarrel sound, not only will the notification module remind other family members, but also the voice change processing module will be used to change the quarrel sound to make it sound smoother, which helps to reduce the intensity of the quarrel and promote family harmony; 3) It can enhance the harmonious atmosphere of the family, through real-time monitoring and identification of family quarrels, and taking corresponding intervention measures, it helps to reduce the occurrence of family quarrels, reduce the negative impact of quarrels on family members, and thus enhance the harmonious atmosphere of the family; 4) It can enhance communication and understanding between family members, and realize information exchange between family members through screen-to-screen linkage technology, which can enhance communication and understanding between each other, and help to timely discover and solve conflicts and problems in the family.

[0085] That is, the technical solution provided by the embodiment of the present invention comprehensively considers various factors, introduces advanced screen-to-screen linkage technology, realizes real-time interaction and information sharing of central control screens between different rooms in the home, and has a built-in sound recognition module in the central control screen, and develops a special quarrel recognition algorithm, which can monitor the sounds in the home environment in real time, accurately identify the quarrel sounds, and after identifying the quarrel sounds, not only reminds other family members through the notification module, but also uses the voice change processing module to change the quarrel sounds to make them sound smoother, so as to reduce the intensity of the quarrel.

[0086] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0087] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0088] According to an embodiment of the present invention, a family quarrel recognition device for implementing the quarrel recognition and processing method is also provided. Figure 4 is a schematic diagram of a device for identifying family quarrels according to an embodiment of the present invention. Figure 4 As shown, the device comprises: an acquisition unit 41, a first processing unit 42 and a second processing unit 43. The following is a description of the device for identifying family quarrels.

[0089] The acquisition unit 41 is used to acquire a sound signal in at least one room in the house, obtain a current sound signal, and identify the current sound signal to determine whether there is a quarrel sound.

[0090] The first processing unit 42 is used to perform voice changing processing on the quarreling sound to obtain voice-changed propagation sound when it is determined that there is a quarreling sound. The voice changing processing is characterized by reducing the speech speed of the quarreling sound and keeping the timbre of the quarreling sound unchanged.

[0091] The second processing unit 43 is used to use multiple central control devices to relay the voice-changing sound, wherein a central control device is installed in each room in the house.

[0092] It should be noted here that the above-mentioned acquisition unit, the first processing unit and the second processing unit correspond to steps S202 to S206 in the above-mentioned embodiment, and the three modules have the same instances and application scenarios implemented by the corresponding steps, but are not limited to the contents disclosed in the above-mentioned embodiment.

[0093] As can be seen from the above, in the scheme recorded in the above-mentioned embodiment of the present invention, first, the acquisition unit can be used to acquire the sound signal in at least one room in the house, obtain the current sound signal, and identify the current sound signal to determine whether there is a quarrel sound; then, the first processing unit can be used to perform voice change processing on the quarrel sound when it is determined that there is a quarrel sound, to obtain a voice-changed propagation sound, and the voice change processing is characterized by reducing the speech speed of the quarrel sound and keeping the timbre of the quarrel sound unchanged; finally, the second processing unit can be used to use multiple central control devices to broadcast the voice-changed propagation sound, wherein a central control device is installed in each room in the house, so as to achieve the screen-to-screen linkage technology, and a sound recognition module is built into the central control screen to accurately identify the quarrel sound, and the notification module is used to remind other family members of the occurrence of the quarrel, and at the same time, the voice change processing module is used to perform voice change processing to reduce the intensity of the quarrel, thereby realizing the information exchange between family members, the technical effect of real-time monitoring and identification of family quarrels, and improving the efficiency and accuracy of communication.

[0094] The technical solution provided by the embodiment of the present invention solves the technical problem that the existing solution uses manual transmission or traditional communication equipment to transmit quarrel information between different rooms, resulting in low transmission efficiency.

[0095] In an optional embodiment, the acquisition unit includes: a preprocessing subunit, used to preprocess the current sound signal to obtain a target sound signal, and perform feature extraction on the target sound signal to obtain feature data, the preprocessing includes deduplication processing and denoising processing; an identification subunit, used to use a machine learning model to identify the feature data to determine whether there is a quarrel sound, the machine learning model includes a support vector machine model, a random forest model and a deep learning model.

[0096] In an optional embodiment, the device for identifying family quarrels further includes: a third processing unit, used to use a machine learning model to process feature data after determining the presence of quarrel sounds, so as to determine a quarrel sound pattern, wherein the quarrel sound pattern represents the intensity of the quarrel; a playback unit, used to use multiple central control devices to play preset music or preset sentences that conform to the quarrel sound pattern, wherein the preset music is music with a rhythm within a preset range, and the preset sentence is a sentence with a speaking speed lower than a preset speaking speed.

[0097] In an optional embodiment, the device for identifying family quarrels further includes: a conversion unit, which is used to obtain sound signals in at least one room in the house, and after obtaining the current sound signal, use speech recognition technology to convert the current sound signal into text data; a preprocessing unit, which is used to preprocess the text data to obtain target text data, and the preprocessing includes deduplication processing and denoising processing; a fourth processing unit, which is used to use natural language processing technology to process the target text data based on at least the syntax of the target text data, the relationship between different speakers, and the emotions of each speaker, and determine the cause of the quarrel.

[0098] In an optional embodiment, the device for identifying family quarrels further includes: a fifth processing unit, which is used to combine natural language processing technology and machine learning models to process current sound signals based on the causes of the quarrel after determining the causes of the quarrel, and generate and use multiple central control devices to play family activity suggestions.

[0099] In an optional embodiment, the device for identifying family quarrels further includes: an optimization unit for optimizing parameters of the natural language processing technology and parameters of the machine learning model by using a K-fold validation method.

[0100] In an optional embodiment, the device for identifying family quarrels further includes: a control unit for controlling the multiple central control devices to display the time, place and intensity of the quarrel in text form after the multiple central control devices are used to relay the voice-changing sound.

[0101] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include a quarrel identification processing method for executing any one of the above items.

[0102] According to another aspect of an embodiment of the present invention, a processor is further provided. The processor is used to run a program. When the program is run, any one of the above-mentioned dispute identification and processing methods is executed.

[0103] According to another aspect of an embodiment of the present invention, a computer program product is provided, including computer instructions, and when the computer instructions are executed by a processor, any one of the above-mentioned dispute identification and processing methods is executed.

[0104] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored program. When the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned dispute identification and processing methods.

[0105] Optionally, in this embodiment, the computer-readable storage medium may be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the communication devices in a communication device group.

[0106] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for executing the following steps: acquiring a sound signal in at least one room in a house, obtaining a current sound signal, and identifying the current sound signal to determine whether there is a quarrel sound; in the case where it is determined that there is a quarrel sound, performing voice changing processing on the quarrel sound to obtain a voice-changed propagation sound, wherein the voice changing processing is characterized by reducing the speech speed of the quarrel sound while keeping the timbre of the quarrel sound unchanged; and using multiple central control devices to broadcast the voice-changed propagation sound, wherein one central control device is installed in each room in the house.

[0107] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: preprocessing the current sound signal to obtain a target sound signal, and performing feature extraction on the target sound signal to obtain feature data, wherein the preprocessing includes deduplication processing and denoising processing; using a machine learning model to identify the feature data to determine whether there is a quarrel sound, wherein the machine learning model includes a support vector machine model, a random forest model, and a deep learning model.

[0108] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: using a machine learning model to process feature data to determine a quarrel sound pattern, wherein the quarrel sound pattern characterizes the intensity of the quarrel; using multiple central control devices to play preset music or preset sentences that conform to the quarrel sound pattern, wherein the preset music is music with a rhythm within a preset range, and the preset sentence is a sentence with a speaking speed lower than a preset speaking speed.

[0109] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: using speech recognition technology to convert the current sound signal into text data; preprocessing the text data to obtain target text data, the preprocessing including deduplication processing and denoising processing; using natural language processing technology, processing the target text data based on at least the syntax of the target text data, the relationship between different speakers, and the emotions of each speaker, to determine the cause of the quarrel.

[0110] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for executing the following steps: based on the cause of the quarrel, combining natural language processing technology and machine learning models to process the current sound signal, generate and use multiple central control devices to play family activity suggestions.

[0111] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: optimizing parameters of the natural language processing technology and parameters of the machine learning model using the K-fold validation method.

[0112] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: controlling a plurality of central control devices to display the time, location and intensity of the quarrel in text form.

[0113] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0114] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the above-mentioned units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0116] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0117] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0118] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the above-mentioned methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0119] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying and processing a quarrel, characterized in that: include: Acquiring a sound signal in at least one room in the house to obtain a current sound signal, and identifying the current sound signal to determine whether there is a quarrel sound; In the case where it is determined that there is a quarrel sound, the quarrel sound is subjected to voice change processing to obtain a voice-changed propagation sound, wherein the voice change processing is characterized by reducing the speech speed of the quarrel sound and keeping the timbre of the quarrel sound unchanged; A plurality of central control devices are used to relay the voice-changing propagation sound, wherein one central control device is respectively installed in each of the rooms in the house.

2. The method according to claim 1, characterized in that Identifying the current sound signal to determine whether there is a quarrel sound includes: Preprocessing the current sound signal to obtain a target sound signal, and extracting features from the target sound signal to obtain feature data, wherein the preprocessing includes deduplication processing and denoising processing; A machine learning model is used to identify the feature data to determine whether the quarrel sound exists, and the machine learning model includes a support vector machine model, a random forest model and a deep learning model.

3. The method according to claim 2, characterized in that After determining that there is a quarrel sound, the method further includes: Processing the characteristic data using a machine learning model to determine a quarrel sound pattern, wherein the quarrel sound pattern represents the intensity of the quarrel; The multiple central control devices are used to play preset music or preset sentences that conform to the quarrel sound pattern, the preset music is music with a rhythm within a preset range, and the preset sentences are sentences with a speaking speed lower than a preset speaking speed.

4. The method according to claim 1, characterized in that After acquiring the sound signal in at least one room in the house and obtaining the current sound signal, the method further includes: Using speech recognition technology to convert the current sound signal into text data; Preprocessing the text data to obtain target text data, wherein the preprocessing includes deduplication processing and denoising processing; The target text data is processed by using natural language processing technology to determine the cause of the quarrel based on at least the syntax of the target text data, the relationship between different speakers and the emotions of each speaker.

5. The method according to claim 4, characterized in that After determining the cause of the quarrel, the method further comprises: Based on the cause of the quarrel, the natural language processing technology and the machine learning model are combined to process the current sound signal, and family activity suggestions are generated and played using the multiple central control devices.

6. The method according to claim 5, characterized in that The method further comprises: The K-fold validation method is used to optimize the parameters of the natural language processing technology and the parameters of the machine learning model.

7. The method according to any one of claims 1 to 6, characterized in that After using multiple central control devices to relay the voice-changing propagation sound, the method further includes: The plurality of central control devices are controlled to display the time, location and intensity of the quarrel in text form.

8. A device for identifying family quarrels, characterized in that: include: an acquisition unit, configured to acquire a sound signal in at least one room in the house, obtain a current sound signal, and identify the current sound signal to determine whether there is a quarrel sound; A first processing unit is configured to, when determining that there is a quarrel sound, perform voice change processing on the quarrel sound to obtain a voice-changed propagation sound, wherein the voice change processing is characterized by reducing the speech speed of the quarrel sound and keeping the timbre of the quarrel sound unchanged; The second processing unit is used to use multiple central control devices to relay the voice-changing sound, wherein one central control device is installed in each of the rooms in the house.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing the dispute identification processing method according to any one of claims 1 to 7.

Images