WO2023184942A1 - Voice interaction method and apparatus and electric appliance - Google Patents

Abstract

A voice interaction method and apparatus, and an electric appliance, the method comprising: receiving a voice input of a target user, and determining target voice information (110); performing voice recognition on the target voice information to obtain a voice recognition result, the voice recognition result comprising at least one of age information of the target user and target emotion features of the target voice information (120); and outputting a response voice, the response voice being set on the basis of the voice recognition result (130). The voice interaction method can make corresponding humanized responses in respect of ages and emotion features according to the voice recognition result, thereby improving the user experience.

Description

Voice interaction method, device and electrical appliance

Cross-references to related applications

This application claims priority to the Chinese patent application with application number 202210324200.7 and titled "Voice Interaction Method, Device and Electrical Appliance" filed on March 29, 2022, which is fully incorporated herein by reference.

Technical field

The present application relates to the technical field of smart electrical appliances, and in particular to a voice interaction method, device and electrical appliances.

Background technique

With the development of science and technology, the application scenarios of voice interaction technology are becoming more and more abundant. By using artificial intelligence and TTS (Text To Speech, from text to speech) and other technologies, voice communication with user groups can be achieved.

In related technologies, the voice response of voice interaction equipment is relatively stiff and cannot provide humanized responses to different users, resulting in a bad user experience for users.

Contents of the invention

The present application provides a voice interaction method, device and electrical appliance to solve the problem of blunt responses in the existing voice interaction devices and achieve the effect of responding emotionally to different users.

This application provides a voice interaction method, including:

Receive the voice input from the target user and determine the target voice information;

Perform speech recognition on the target speech information to obtain a speech recognition result; the speech recognition result includes at least one of the age information of the target user and the target emotional characteristics of the target speech information;

A response voice is output, and the response voice is set based on the voice recognition result.

According to a voice interaction method provided by this application, performing voice recognition on the target voice information to obtain a voice recognition result includes:

Obtain the voiceprint feature recognition results related to the target voice information;

Based on the voiceprint feature recognition result, determine the age information of the target user;

Based on the age information, a first emotional characteristic of the response voice is set.

According to a voice interaction method provided by this application, obtaining the voiceprint feature recognition results related to the target voice information includes:

Based on the target voice information, determine the target voiceprint characteristics;

Search for voiceprint feature samples that match the target voiceprint feature in a voiceprint feature database, where a voiceprint feature sample set is pre-stored in the voiceprint feature database;

When a voiceprint feature sample matching the target voiceprint feature is found, age information corresponding to the voiceprint feature sample is determined.

According to a voice interaction method provided by this application, after searching for a voiceprint feature sample that matches the target voiceprint feature in the voiceprint feature database, the method further includes:

When no voiceprint feature sample matching the target voiceprint feature is found, the target voiceprint feature is input to the age prediction model to obtain the age information output by the age prediction model, and the age prediction The model is trained using age sample voiceprint features as samples and the age information of the speaker in the age sample voiceprint features as labels.

According to a voice interaction method provided by this application, after obtaining the voiceprint feature recognition results related to the target voice information, the method further includes:

If the age information of the target user is not determined based on the voiceprint feature recognition result, perform voice emotion recognition on the target voice information to obtain the target emotion feature of the target voice information;

Based on the target emotional feature, a second emotional feature of the response voice is set.

According to a voice interaction method provided by this application, performing voice emotion recognition on the target voice information to obtain the target emotion characteristics of the target voice information includes:

Based on the target voice information, generate target text information;

Extract target corpus from the target text information;

When the target corpus is found in the emotional dictionary, determine the emotional characteristics of the corpus corresponding to the target corpus;

Based on the emotional characteristics of the corpus, determine the target emotional characteristics;

The emotion dictionary includes a plurality of corpus and emotional features corresponding to the corpus.

This application also provides a voice interaction device, including:

The receiving module is used to receive the voice input of the target user and determine the target voice information;

A processing module configured to perform speech recognition on the target speech information and obtain a speech recognition result; the speech recognition result includes at least one of the age information of the target user and the target emotional characteristics of the target speech information;

An output module is configured to output a response voice, where the response voice is set based on the voice recognition result.

This application also provides an electrical appliance, including the above-mentioned voice interaction device.

The application also provides an air conditioner, including an indoor unit, an outdoor unit, a processor and a memory provided in the indoor unit or outdoor unit; and a program stored on the memory and executable on the processor. Or an instruction, when the program or instruction is executed by the processor, any one of the above voice interaction methods is executed.

This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements any one of the above voice interaction methods. .

This application also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements any one of the above voice interaction methods.

The present application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements any one of the above voice interaction methods.

The voice interaction method, device and electrical appliance provided by this application can obtain the age of the target user or the target emotional characteristics of the target voice information by performing speech recognition on the target voice information of the target user, and then can make decisions based on the age and emotional characteristics based on the speech recognition results. Provide corresponding humanized responses, thereby improving user experience.

Description of drawings

In order to explain the technical solutions in this application or the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are of the present invention. For some embodiments of the application, those of ordinary skill in the art can also obtain other drawings based on these drawings without exerting creative efforts.

Figure 1 is one of the flow diagrams of the voice interaction method provided by this application;

Figure 2 is the second schematic flow chart of the voice interaction method provided by this application;

Figure 3 is a schematic structural diagram of the voice interaction device provided by this application;

Figure 4 is a schematic structural diagram of an electronic device provided by this application.

Detailed ways

In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application. , not all examples. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

The voice interaction method, device and electrical appliance of the present application will be described below with reference to Figures 1-4.

The execution subject of the voice interaction method in the embodiment of the present application may be a processor. Of course, in some embodiments, the execution subject of the voice interaction method in the embodiment of the present application may also be a server, which is not limited here. The voice interaction method in the embodiment of the present application will be described below by taking the execution subject as a processor as an example.

As shown in Figure 1, the voice interaction method in the embodiment of the present application mainly includes step 110, step 120 and step 130.

Step 110: Receive the target user's voice input and determine the target voice information.

It should be noted that after the target user sends a voice, the voice input can be received by collecting the target user's voice.

In some embodiments, the target user's voice can be collected through devices such as pickups or microphones. Of course, in other embodiments, other voice collection devices can also be used to collect the target user's voice.

In a home scenario, when the target user's voice is collected through a microphone, the microphone can be integrated with electrical equipment. For example, the pickup can be installed on the indoor unit housing of the air conditioner. In this case, voice interaction with the target user can be achieved through air conditioning.

Alternatively, when the target user's voice is collected through a microphone, the microphone may be a microphone configured on the target user's smartphone. In this case, voice interaction with the target user can be achieved through the mobile phone.

Of course, in some other scenarios, the pickup or microphone can also be installed at other locations. There is no restriction on the medium of voice interaction here, and there is no restriction on the installation position and installation method of the pickup or microphone.

In some embodiments, the wake-up of voice reception can be achieved through a special wake-up voice, that is, when the wake-up voice is received, the voice of the target user is collected and the target voice information is determined.

For example, when the collected voice is "Xiaoyou, Xiaoyou, turn on the air conditioner!", "Xiaoyou" can be recognized as the wake-up voice. In this case, the wake-up voice is recognized before starting to collect the target user's voice, thereby realizing the reception of the target user's voice input.

Of course, in some other embodiments, environmental sounds can also be collected continuously, and after the target user's voice is recognized, the target voice information is determined.

It can be understood that after collecting the target user's voice, simple pre-processing can be performed on the collected voice, such as voice separation and noise reduction, to obtain an effective voice information flow, and then obtain effective target voice information. .

Step 120: Perform speech recognition on the target speech information to obtain a speech recognition result.

It can be understood that the speech recognition result at least includes text information in the target speech information.

It should be noted that the speech recognition result also includes at least one of the target user's age information and the target emotional characteristics of the target voice information.

It can be understood that by performing speech recognition on the target speech information, the age information of the target user corresponding to the target speech information can be determined.

The age information can be a specific age value or a certain age range. In this embodiment, the target user can be determined to be a child, a middle-aged person, an elderly person, etc. based on the age information.

For example, when the value of the determined age information is in the interval 0-14, the target user can be determined to be a child; when the value of the determined age information is greater than 55, the target user can be determined to be an elderly person.

In some embodiments, the age information of the target user corresponding to the target voice information can be determined by extracting voiceprint features of the target voice information.

In other embodiments, the age information of the target user may be determined by matching the target voice information with user voices in a preset user library.

Of course, in other implementations, the age information of the target user corresponding to the target voice information can also be determined in other ways, which is not limited here.

It can be understood that by performing speech recognition on the target speech information, the target emotional characteristics of the target speech information can be determined.

The target emotional features are used to represent the emotional color of the target voice information of the target object. According to different classification standards, the target emotional features can be divided into different types.

In some embodiments, the target emotional characteristics can be divided into positive emotions, neutral emotions, and negative emotions according to positive or negative factors.

Positive emotions can include emotions such as pleasure, joy, and relaxation, negative emotions can include emotions such as pain, frustration, sadness, and anger, and neutral emotions can include emotions such as seriousness, determination, and doubt.

It can be understood that a database can be constructed based on the tone information and text information corresponding to positive emotions, neutral emotions and negative emotions, and the database contains different emotional characteristics corresponding to different tone information and text information.

In this embodiment, the target emotional characteristics of the target voice information can be determined based on the tone information, text information and database in the target voice information.

For example, when the content of the target voice message is "Xiaoyou, Xiaoyou, I'm so annoyed!", the voice message contains the negative information "annoying", so the target emotional characteristic of the target voice message can be determined as negative emotion.

In some embodiments, the target speech information can be input into the speech emotion recognition model, and the target speech information can be comprehensively recognized and analyzed to obtain the target emotion characteristics of the target speech information.

In other embodiments, the emotional words can be matched by extracting relevant corpus in the target speech information, thereby obtaining the target emotional characteristics of the target speech information.

Of course, in other implementations, the target emotional characteristics of the target speech information can also be determined through other methods, which are not limited here.

Step 130: Output a response voice, which is set based on the voice recognition result.

After obtaining the recognition result of the target voice information, a response voice can be obtained for the target voice information.

It should be noted that the response voice includes voice broadcast content that responds to the text information in the target voice information, and the broadcast voice can be generated based on TTS technology.

Since the speech recognition results include the age information of the target user and the target emotional characteristics of the target voice information, the response voice can be adaptively set according to the age of the target user and the target emotional characteristics of the voice interaction when broadcasting, thereby achieving The effect of personalized voice interaction for different types of target users.

According to the voice interaction method of the embodiment of the present application, by performing speech recognition on the target voice information of the target user, the age of the target user or the target emotional characteristics of the target voice information can be obtained, and then the corresponding age and emotional characteristics can be made according to the speech recognition results. humanized responses, thereby improving user experience.

As shown in Figure 2, in some embodiments, step 120: Perform speech recognition on the target speech information to obtain a speech recognition result, which mainly includes step 1201, step 1202 and step 1203.

Step 1201: Obtain the voiceprint feature recognition results related to the target voice information.

It can be understood that the voiceprint feature recognition of the target voice information may include identifying the identity, age, gender and other information of the target user corresponding to the target voice information.

Step 1202: Determine the age information of the target user based on the voiceprint feature recognition result.

In some embodiments, the age information of the target user may be determined based on the identified identity information of the target user.

In other embodiments, the age information of the target user can be determined directly through the voiceprint feature recognition results.

In this case, step 1201: Obtaining the voiceprint feature recognition result related to the target voice information may specifically include determining the target voiceprint feature based on the target voice information.

It can be understood that before performing voiceprint feature recognition on the target voice information, it is necessary to extract the voiceprint features of the target voice information.

In some embodiments, voiceprint features can be extracted from the target voice information through a pre-trained voiceprint feature extraction neural network model.

It can be understood that the pre-trained neural network model can be trained using the sample voice information as the sample and the voiceprint characteristics of the speaker in the sample voice information as the label.

The specific training process can be as follows: input the sample voice information into the voiceprint feature extraction neural network model, output the identified voiceprint feature of the speaker, and use the similarity between the identified voiceprint feature and the label as the loss. According to the loss Adjust the amount of updates needed in the voiceprint feature extraction neural network model until the loss is less than the preset threshold or the number of training tests reaches the preset number.

After extracting the target voiceprint feature in the target voice information, a voiceprint feature sample matching the target voiceprint feature is searched for in the voiceprint feature database.

It can be understood that a voiceprint feature sample set is pre-stored in the voiceprint feature database, and the voiceprint feature sample set includes each voiceprint feature sample and the age information corresponding to each voiceprint feature sample.

During the search process, the matching degree between the target voiceprint feature and each voiceprint feature sample can be calculated.

When calculating the matching degree, the matching degree between the target voiceprint feature and each voiceprint feature sample can be calculated through a linear discriminant model. When the matching degree is greater than the preset value, the voiceprint feature sample will be The corresponding age information is used as the age information corresponding to the target voiceprint feature.

Of course, in some embodiments, when the matching degree between the target voiceprint feature and multiple voiceprint feature samples is greater than the preset value, the voiceprint feature sample with the largest matching degree among the multiple voiceprint feature samples that meets the requirements can be The corresponding age information is used as the age information corresponding to the target voiceprint feature.

In other words, when a voiceprint feature sample matching the target voiceprint feature is found, the age information corresponding to the voiceprint feature sample can be determined, and then the age information corresponding to the voiceprint feature sample is used as the target voiceprint feature. age information.

According to the voice interaction method of the embodiment of the present application, by searching the target voiceprint feature in the voiceprint feature database, the age information corresponding to the target voiceprint feature can be quickly and accurately confirmed, thereby improving the accuracy of the voice interaction response content. , thereby improving the user experience.

In some embodiments, after searching for voiceprint feature samples that match the target voiceprint feature in the voiceprint feature database, the voice interaction method in the embodiment of the present application further includes:

When no voiceprint feature sample matching the target voiceprint feature is found, the target voiceprint feature is input to the age prediction model to obtain the age information output by the age prediction model.

It can be understood that the age prediction model is trained using age sample voiceprint features as samples and the age information of the speaker in the age sample voiceprint features as labels.

It can be understood that the age prediction model can be a convolutional neural network model, a hidden Markov model or a Gaussian mixture model, etc. There is no restriction on the type of the age prediction model here.

In this embodiment, after no voiceprint feature sample matching the target voiceprint feature is found in the voiceprint feature database, the age prediction value of the target information corresponding to the target voice information can be obtained through the age prediction model, which can accurately Confirming the age information corresponding to the target voiceprint characteristics improves the accuracy of the voice interactive response content, thereby improving the user experience.

Step 1203: Based on the age information, set the first emotional feature of the response voice.

It is understandable that according to the different age information of the target users, the first emotional feature of the response voice can be set to achieve a more humanized voice interaction.

In some embodiments, when the value of the determined age information is in the interval 0-14, the target user can be determined to be a child; when the value of the determined age information is greater than 55, the target user can be determined to be an elderly person; The value of the determined age information is between 14 and 55, and the target user can be determined to be a middle-aged person.

According to the determined age information, when it is determined that the target user corresponding to the target voice information is a child, the first emotional feature of the response voice may be a positive emotional feature. Positive emotions can include emotions such as joy, joy, liveliness, and relaxation.

In this case, the response voice can be set to a more pleasant, lively and relaxing tone, and more pleasant and relaxing words can be added to the content of the response voice, so as to better meet the needs of children.

According to the determined age information, when it is determined that the target user corresponding to the target voice information is an elderly person, the first emotional feature of the response voice may be a neutral emotional feature. Neutral emotions can include emotions such as seriousness and firmness.

In this case, the response voice can be set to a more serious and firm tone, and more formal and affirmative vocabulary can be added to the content of the response voice, so as to better meet the needs of the elderly.

According to the determined age information, when it is determined that the target user corresponding to the target voice information is a middle-aged person, the first emotional feature of the response voice may be a neutral emotional feature or a positive emotional feature.

In this case, the response voice can be set to a more serious or pleasant tone, and more formal and pleasant vocabulary can be added to the content of the response voice, so as to better meet the needs of middle-aged people.

Of course, in other embodiments, other types of emotional features can also be set for target users of different ages. Target users of different ages can correspond to other different types of emotional features, which can be set according to the actual situation. Here, different types of emotional features can be set for target users of different ages. There are no restrictions on the types of emotional characteristics corresponding to target users in age groups.

According to the voice interaction method of the embodiment of the present application, by determining the age information of the target user corresponding to the target voice information, and then setting the emotional characteristics of the output response voice based on the age information, the emotional characteristics of the response voice can be adjusted for target users of different ages. Settings on the phone make the voice response more humane, meet the different personalized needs of different users, and improve the user experience.

In some embodiments, in step 1201: after obtaining the voiceprint feature recognition results related to the target voice information, the voice interaction method in the embodiment of the present application also includes:

When the age information of the target user is not determined based on the voiceprint feature recognition results, speech emotion recognition is performed on the target voice information to obtain the target emotion characteristics of the target voice information.

It is understandable that when identifying target voiceprint features, if the age information of the target user cannot be determined from the voiceprint feature recognition results, the target emotional features of the target voice information can be obtained by performing voice emotion recognition on the target voice information. .

In some embodiments, a voiceprint feature sample whose matching degree with the target voiceprint feature meets the preset conditions cannot be found in the voiceprint feature database, and the age prediction model cannot output age information corresponding to the target voiceprint feature. , speech emotion recognition can be performed on the target speech information.

In some embodiments, speech emotion recognition can be performed on the target speech information through a speech emotion recognition neural network model.

It can be understood that the speech emotion recognition neural network model is trained by taking speech information with emotional characteristics as samples and using the emotional characteristics of the speech information with emotional characteristics as labels.

In other embodiments, performing speech emotion recognition on the target speech information and obtaining the target emotion characteristics of the target speech information specifically includes generating target text information based on the target speech information.

It is understandable that in the process of speech emotion recognition, the target speech information is first converted into text information to obtain the target text information.

After obtaining the target text information, extract the target corpus in the target text information. It can be understood that the target corpus is the words or words with emotional tendencies in the target text information.

For example, when the target text information of the target speech information is "Xiaoyou, Xiaoyou, I'm so annoyed!", the target corpus can be "annoyed" and "good". "Annoying" is an emotional word, which can express the target user's negative emotions, and "good" is an adverb of degree, which can be used to express the degree of the target user's negative emotions.

When the target corpus is found in the emotional dictionary, the emotional characteristics of the corpus corresponding to the target corpus are determined.

It can be understood that the emotional dictionary includes multiple corpora and emotional features corresponding to the corpora.

After determining the corpus emotional characteristics corresponding to the target corpus, the target emotional characteristics can be determined based on the corpus emotional characteristics.

In this implementation, the corresponding emotional features can be searched in the emotional dictionary for the target corpus. When there is only a single target corpus in the target text information, the emotional features of the target corpus queried in the emotional dictionary can be directly used as the target emotional features.

When there are multiple target corpora in the target text information, corresponding weight processing is performed on the emotional features corresponding to the multiple target corpora.

For example, different weight scores can be set for different emotional features corresponding to different target corpus. In some embodiments, for different corpus, different corpus can be divided into positive words, negative words, negative words, adverbs of degree, etc.

In this case, you can add the weight scores of positive words, subtract the weight scores of negative words, take the inverse of the weight of negative words, and multiply the weight of the adverb of degree and the weight of the words it modifies to get the final result. Weight score.

According to the final weight score, the emotional characteristics can be determined to be a certain type of positive emotion, neutral emotion, or negative emotion.

According to the voice interaction method of the embodiment of the present application, by obtaining the target text information of the target speech information, and then analyzing the target corpus in the target text information, the corpus emotional characteristics of the target corpus are obtained, and then the target emotional characteristics of the target speech information are determined. , can more accurately determine the target emotional characteristics of the target voice information, and then facilitate the setting of corresponding response voices for target users with different emotional characteristics.

It can be understood that after the target emotional characteristics of the target voice information are determined, the second emotional characteristics of the response voice can be set based on the target emotional characteristics.

In some embodiments, when it is determined that the target emotional feature is a positive emotional feature, the second emotional feature of the response voice may be a positive emotional feature. Positive emotions can include emotions such as joy, joy, liveliness, and relaxation.

For example, the target emotional characteristics include happy and relaxed emotions. The response voice can be set to a more pleasant, lively and relaxed tone. More pleasant and relaxing words can be added to the content of the response voice, which can better fit the current needs of the target users. Feeling.

In some embodiments, when it is determined that the target emotional feature is a negative emotional feature, the second emotional feature of the response voice may be a neutral emotional feature. Negative emotions can include emotions such as anger and impatience.

For example, if the target emotional characteristics include angry emotions, the response voice can be set to a more serious and affirmative tone, and more formal and affirmative words can be added to the content of the response voice to better fit the current mood of the target user.

Of course, in other embodiments, other types of second emotional features can also be set for different target emotional features. The second emotional features can be set according to the actual situation. The type of the second emotional feature is not limited here.

According to the voice interaction method of the embodiment of the present application, by determining the target emotional characteristics corresponding to the target voice information, and then setting different second emotional characteristics according to the target emotional characteristics to set the response voice, it is possible to emotionally respond to the emotional characteristics of different target users. Setting the response voice makes the response voice more humane, more suitable for the different personalized needs of different users, and improves the user experience.

As shown in FIG. 3 , the interaction device provided by the present application is described below. The voice interaction device described below and the voice interaction method described above can be mutually referenced.

The voice interaction device in this embodiment of the present application includes a receiving module 310, a processing module 320, and an output module 330.

The receiving module 310 is used to receive the voice input of the target user and determine the target voice information;

The processing module 320 is configured to perform speech recognition on the target speech information and obtain a speech recognition result; the speech recognition result includes at least one of the age information of the target user and the target emotional characteristics of the target speech information;

The output module 330 is used to output the response voice, and the response voice is set based on the voice recognition result.

According to the voice interaction device provided by the embodiment of the present application, by performing speech recognition on the target speech information of the target user, the age of the target user or the target emotional characteristics of the target speech information can be obtained, and then the target user's age and emotional characteristics can be determined based on the speech recognition results. Provide corresponding humanized responses, thereby improving user experience.

In some embodiments, the processing module 320 is also used to obtain the voiceprint feature recognition results related to the target voice information; determine the age information of the target user based on the voiceprint feature recognition results; and set the first emotional feature of the response voice based on the age information. .

In some embodiments, the processing module 320 is also used to determine the target voiceprint features based on the target voice information; search for voiceprint feature samples that match the target voiceprint features in the voiceprint feature database, and pre-set the target voiceprint features in the voiceprint feature database. A voiceprint feature sample set is stored; when a voiceprint feature sample matching the target voiceprint feature is found, the age information corresponding to the voiceprint feature sample is determined.

In some embodiments, the processing module 320 is also configured to input the target voiceprint feature to the age prediction model when no voiceprint feature sample matching the target voiceprint feature is found, and obtain the age output by the age prediction model. Information, the age prediction model is trained using age sample voiceprint features as samples and the age information of the speaker in the age sample voiceprint features as labels.

In some embodiments, the processing module 320 is also used to perform speech emotion recognition on the target voice information when the age information of the target user is not determined based on the voiceprint feature recognition result, and obtain the target emotion characteristics of the target voice information; Based on the target emotional characteristics, set the second emotional characteristics of the response voice.

In some embodiments, the processing module 320 is also used to generate target text information based on the target speech information; extract the target corpus in the target text information; and determine the corpus corresponding to the target corpus when the target corpus is found in the emotional dictionary. Emotional features; based on the emotional features of the corpus, determine the target emotional features; the emotional dictionary includes multiple corpora and the emotional features corresponding to the corpora.

An embodiment of the present application also provides an electrical appliance. The electrical appliance includes the above-mentioned voice interaction device and can perform voice interaction with the user. Electrical appliances can be air conditioners, televisions, washing machines, refrigerators, water purifiers, etc. There is no restriction on the type of electrical appliances.

Embodiments of the present application also provide an air conditioner, which includes an indoor unit, an outdoor unit, a processor and a processor provided in the indoor unit or the outdoor unit; and also includes a program or instructions stored in a memory and executable on the processor, When the program or instruction is executed by the processor, the voice interaction method as described above is performed.

Figure 4 illustrates a schematic diagram of the physical structure of an electronic device. As shown in Figure 4, the electronic device may include: a processor (processor) 410, a communications interface (Communications Interface) 420, a memory (memory) 430 and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 complete communication with each other through the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute the voice interaction method. The method includes: receiving voice input from the target user and determining the target voice information; performing voice recognition on the target voice information to obtain a voice recognition result; the voice recognition result At least one of the target emotional characteristics including age information of the target user and target voice information; outputting a response voice, where the response voice is set based on the voice recognition result.

In addition, the above-mentioned logical instructions in the memory 430 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

On the other hand, the present application also provides a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can Execute the voice interaction method provided by each of the above methods. The method includes: receiving the voice input of the target user and determining the target voice information; performing voice recognition on the target voice information to obtain a voice recognition result; the voice recognition result includes the age information of the target user and At least one of the target emotional characteristics of the target voice information; output a response voice, and the response voice is set based on the speech recognition result.

On the other hand, the present application also provides a non-transitory computer-readable storage medium on which a computer program is stored. The computer program is implemented when executed by a processor to execute the voice interaction method provided by each of the above methods. The method includes: Receive the target user's voice input and determine the target voice information; perform voice recognition on the target voice information to obtain a voice recognition result; the voice recognition result includes at least one of the target user's age information and the target emotional characteristics of the target voice information; and output a response voice. , the response voice is set based on the speech recognition results.

The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the part of the above technical solution that essentially contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., including a number of instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application, but not to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions in the embodiments of the present application.

Claims (10)

A voice interaction method including: Receive the voice input from the target user and determine the target voice information; Perform speech recognition on the target speech information to obtain a speech recognition result; the speech recognition result includes at least one of the age information of the target user and the target emotional characteristics of the target speech information; A response voice is output, and the response voice is set based on the voice recognition result. The voice interaction method according to claim 1, wherein said performing voice recognition on the target voice information to obtain a voice recognition result includes: Obtain the voiceprint feature recognition results related to the target voice information; Based on the voiceprint feature recognition result, determine the age information of the target user; Based on the age information, a first emotional characteristic of the response voice is set. The voice interaction method according to claim 2, wherein said obtaining the voiceprint feature recognition results related to the target voice information includes: Based on the target voice information, determine the target voiceprint characteristics; Search for voiceprint feature samples that match the target voiceprint feature in a voiceprint feature database, where a voiceprint feature sample set is pre-stored in the voiceprint feature database; When a voiceprint feature sample matching the target voiceprint feature is found, age information corresponding to the voiceprint feature sample is determined. The voice interaction method according to claim 3, wherein after searching for voiceprint feature samples matching the target voiceprint feature in the voiceprint feature database, the method further includes: When no voiceprint feature sample matching the target voiceprint feature is found, the target voiceprint feature is input to the age prediction model to obtain the age information output by the age prediction model, and the age prediction The model is trained using age sample voiceprint features as samples and the age information of the speaker in the age sample voiceprint features as labels. The voice interaction method according to any one of claims 2 to 4, wherein after obtaining the voiceprint feature recognition result related to the target voice information, the method further includes: If the age information of the target user is not determined based on the voiceprint feature recognition result, perform voice emotion recognition on the target voice information to obtain the target emotion feature of the target voice information; Based on the target emotional feature, a second emotional feature of the response voice is set. The voice interaction method according to claim 5, wherein the performing voice emotion recognition on the target voice information to obtain the target emotion characteristics of the target voice information includes: Based on the target voice information, generate target text information; Extract target corpus from the target text information; When the target corpus is found in the emotional dictionary, determine the emotional characteristics of the corpus corresponding to the target corpus; Based on the emotional characteristics of the corpus, determine the target emotional characteristics; The emotion dictionary includes a plurality of corpus and emotional features corresponding to the corpus. A voice interaction device including: The receiving module is used to receive the voice input of the target user and determine the target voice information; A processing module configured to perform speech recognition on the target speech information and obtain a speech recognition result; the speech recognition result includes at least one of the age information of the target user and the target emotional characteristics of the target speech information; An output module is configured to output a response voice, where the response voice is set based on the voice recognition result. An electrical appliance including the voice interaction device according to claim 7. An air conditioner includes an indoor unit, an outdoor unit, a processor and a memory provided in the indoor unit or outdoor unit; and also includes a program or instructions stored on the memory and executable on the processor, wherein , when the program or instruction is executed by the processor, the voice interaction method according to any one of claims 1 to 6 is executed. An electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, any one of claims 1 to 6 is implemented. The voice interaction method described in the item.

Images