Voice interaction system and voice interaction method

Abstract

The invention relates to the technical field of voice interaction, and discloses a voice interaction system and a voice interaction method.The voice interaction system comprises a voice acquisition module, a voice recognition module, a semantic understanding module, a response generation module, an output module and a control module; the voice acquisition module is used for acquiring voice input of a user and converting an acquired voice signal into a digital signal; and the voice recognition module is matched with the voice acquisition module and is used for receiving the digital signal, performing voice recognition processing on the digital signal and converting the digital signal into text information. According to the system, a knowledge base sub-module of the semantic understanding module is used for storing domain knowledge related to user intentions and key information and providing rich background information for semantic understanding and intention recognition, and the third main control module is responsible for controlling the semantic understanding process and inputting text information into a knowledge base for comparison and analysis; the intention and the demand of the user are determined, and the adaptation degree to the user is improved.

Description

Technical Field

[0001] The present invention relates to the field of voice interaction technology, and in particular to a voice interaction system and a voice interaction method. Background Art

[0002] The speaker is the terminal of the entire sound system. Its function is to convert audio electrical energy into corresponding sound energy and radiate it into space. It is an extremely important component of the sound system because it is responsible for the key task of converting electrical signals into sound signals for people to listen directly. It can be considered that the performance of the speaker plays a key role in the playback quality of a sound system. When a general speaker is connected to a power source and external devices such as mobile phones, computers, MP3s, etc., and audio data is transmitted to it, the speaker can play according to the data.

[0003] After searching, the Chinese patent number CN110503943B discloses a voice interaction method and a voice interaction system. The method includes: a preprocessing step, preprocessing the input voice information and outputting the voice segment; a semantic recognition step, semantic recognition of the voice segment output by the preprocessing step and outputting semantic information; a gender classification step, identifying the user's gender for the voice segment output by the preprocessing step and outputting gender information; and a fusion processing step, fusing the gender information and the semantic information to obtain personalized reply information for the voice information. According to the voice interaction method and voice interaction system of the present invention, it is possible to differentiate replies according to the user's gender, improve user experience, and improve the intelligence of voice interaction.

[0004] After searching, the Chinese patent number CN110503943A discloses a voice interaction method and a voice interaction system. The method includes: a preprocessing step, preprocessing the input voice information and outputting the voice segment; a semantic recognition step, semantic recognition of the voice segment output by the preprocessing step and outputting semantic information; a gender classification step, identifying the user's gender for the voice segment output by the preprocessing step and outputting gender information; and a fusion processing step, fusing the gender information and the semantic information to obtain personalized reply information for the voice information. According to the voice interaction method and voice interaction system of the present invention, it is possible to differentiate replies according to the user's gender, improve user experience, and improve the intelligence of voice interaction.

[0005] However, the existing voice interaction system does not support the personalized needs of users. Each user has his or her own language habits and expressions, and the traditional voice interaction system can only recognize and process standardized voice commands, which is difficult to meet the personalized needs of users, limiting the application scenarios and user experience of the voice interaction system. Based on this, the present invention designs a voice interaction system and a voice interaction method to solve the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide a voice interaction system and a voice interaction method, which solve the problem that personalization is difficult to satisfy in the background technology.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a voice interaction system, including a voice acquisition module, a voice recognition module, a semantic understanding module, a response generation module, an output module and a control module;

[0008] The voice acquisition module is used to acquire the user's voice input and convert the acquired voice signal into a digital signal;

[0009] The speech recognition module cooperates with the speech acquisition module to receive digital signals, perform speech recognition processing on the digital signals, and convert the digital signals into text information;

[0010] The semantic understanding module cooperates with the speech recognition module to receive text information, perform semantic analysis and understanding on the text information, and determine the user's intention and needs;

[0011] The response generation module is connected to the semantic understanding module to generate corresponding response text or voice signal according to the user's intention and needs;

[0012] The output module cooperates with the response generation module to output the response text or voice signal to the user;

[0013] The control module is connected to the other modules and is used to coordinate and control the data interaction and function execution between the modules.

[0014] A voice interaction method of a voice interaction system,

[0015] Step S1: The voice acquisition module acquires the user's voice input, converts the acquired voice signal into a digital signal, and inputs the digital signal into the voice recognition module for voice recognition processing;

[0016] Step S2: After the speech recognition module processes the received digital signal, it converts the speech signal into text information to facilitate subsequent analysis and understanding;

[0017] Step S3: inputting the text information into the semantic understanding module for semantic analysis and understanding to determine the user's intentions and needs;

[0018] Step S4: generating a corresponding response text or voice signal through the response generation module according to the user's intention and needs;

[0019] Step S5: outputting the response text or voice signal to the user through the output module. The control module is connected with each module to coordinate and control the data interaction and function execution between each module.

[0020] Preferably, the voice collection module includes a sound pickup module and a first main control module, the sound pickup module includes a plurality of microphones arranged in an array for receiving sound; the voice recognition module also includes a USB access determination module.

[0021] Preferably, the speech recognition module includes a CPU control processing module and a second main control module, adopts a deep learning algorithm to perform speech recognition processing, and can recognize speech signals in multiple languages ​​and accents.

[0022] Preferably, the semantic understanding module includes a knowledge base submodule and a third main control module, which are used to store domain knowledge related to user intentions and key information to assist the semantic understanding module in intent recognition and key information extraction, and the third main control module is electrically connected to the CPU control processing module; the speech recognition processing also includes gender classification processing and unified processing, which identifies the user's gender and outputs gender information, and uniformly processes gender information and semantic information to obtain personalized response information for the voice information.

[0023] Preferably, the response generation module includes a personalized response submodule for generating personalized response content according to the user's personal preferences and historical interaction records; the output module includes an audio port and a speaker for final transmission of sound.

[0024] Preferably, the control module is electrically connected to the CPU control processing module, and the control module also includes a filter module for extracting and constructing the output acoustic features of the filter, and inputting the output acoustic features of the filter and the file into the overall model for model training until the model converges.

[0025] Preferably, in step S2, the speech recognition module further includes a noise suppression submodule for performing noise suppression processing on the digital signal to improve the accuracy of speech recognition.

[0026] Preferably, the step S3 further includes utilizing the domain knowledge in the knowledge base submodule to assist the semantic understanding module in performing intent recognition and key information extraction.

[0027] Preferably, the step S4 further includes using a personalized response submodule to generate personalized response content according to the user's personal preferences and historical interaction records.

[0028] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0029] 1. In the present invention, the knowledge base submodule of the semantic understanding module is used to store domain knowledge related to user intentions and key information, providing rich background information for semantic understanding and intention recognition, while the third main control module is responsible for controlling the semantic understanding process, inputting text information into the knowledge base for comparison and analysis, determining the user's intentions and needs, improving the adaptability to the user, and meeting the user's personalized needs.

[0030] 2. In the present invention, the CPU control processing module of the speech recognition module serves as the core processor of the module and is responsible for executing the speech recognition algorithm and converting the digital signal into text information. The second main control module works in conjunction with the CPU control processing module to manage the entire recognition process, including the reception, processing and output of data. The noise suppression submodule performs noise suppression processing on the digital signal to remove interference factors such as background noise and echo, thereby improving the accuracy of speech recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a work flow chart of the present invention;

[0032] Figure 2 It is a system structure diagram of the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0034] Embodiment 1;

[0035] See also Figure 1-Figure 2 , a voice interaction system, including a voice acquisition module, a voice recognition module, a semantic understanding module, a response generation module, an output module and a control module,

[0036] The voice acquisition module is used to collect the user's voice input and convert the collected voice signal into a digital signal;

[0037] The speech recognition module cooperates with the speech acquisition module to receive digital signals, perform speech recognition processing on the digital signals, and convert the digital signals into text information;

[0038] The semantic understanding module cooperates with the speech recognition module to receive text information, perform semantic analysis and understanding on the text information, and determine the user's intention and needs;

[0039] The response generation module is connected to the semantic understanding module to generate corresponding response text or voice signals according to the user's intention and needs;

[0040] The output module cooperates with the response generation module to output the response text or voice signal to the user;

[0041] The control module is connected to other modules to coordinate and control data interaction and function execution between modules.

[0042] like Figure 2 As shown, the voice collection module includes a sound pickup module and a first main control module. The sound pickup module includes a plurality of microphones arranged in an array for receiving sound. The voice recognition module also includes a USB access judgment module.

[0043] like Figure 2 As shown, the speech recognition module includes a CPU control processing module and a second main control module, which uses a deep learning algorithm for speech recognition processing and can recognize speech signals in multiple languages ​​and accents.

[0044] like Figure 2 As shown, the semantic understanding module includes a knowledge base submodule and a third main control module, which are used to store domain knowledge related to user intentions and key information, and assist the semantic understanding module in intention recognition and key information extraction. The third main control module is electrically connected to the CPU control processing module; the speech recognition processing also includes gender classification processing and unified processing, which identifies the user's gender and outputs gender information, and uniformly processes gender information and semantic information to obtain personalized response information for the voice information.

[0045] like Figure 2 As shown, the response generation module includes a personalized response submodule, which is used to generate personalized response content according to the user's personal preferences and historical interaction records; the output module includes an audio port and a speaker for the final transmission of sound.

[0046] like Figure 2 As shown, the control module is electrically connected to the CPU control processing module. The control module also includes a filter module for extracting and constructing the output acoustic features of the filter, and inputting the output acoustic features of the filter and the file into the overall model for model training until the model converges.

[0047] The working principle of the embodiment of the present invention is as follows: the voice collection module includes a sound pickup module and a first main control module. The sound pickup module is used to receive the user's voice input in all directions through a plurality of microphones arranged in an array. This array design not only enhances the receiving capability of voice signals, but also improves the positioning and separation capability of voices through signal processing technology, which helps to identify the target voice in a complex environment. The first main control module is responsible for controlling the operation of the sound pickup module and converting the collected voice signals into digital signals. The conversion process may also involve pre-processing steps such as signal enhancement and filtering to improve the accuracy of subsequent voice recognition processing.

[0048] The CPU control processing module of the speech recognition module is the core processor of the module, responsible for executing the speech recognition algorithm and converting digital signals into text information. The second main control module works in conjunction with the CPU control processing module to manage the entire recognition process, including data reception, processing and output. USB access judgment module: used to detect whether an external device (such as a USB microphone) is connected to the system, so as to flexibly adapt to different input devices. The noise suppression submodule performs noise suppression processing on the digital signal to remove interference factors such as background noise and echo, and improve the accuracy of speech recognition.

[0049] The knowledge base submodule of the semantic understanding module is used to store domain knowledge related to user intentions and key information, providing rich background information for semantic understanding and intention recognition, while the third main control module is responsible for controlling the semantic understanding process, inputting text information into the knowledge base for comparison and analysis, and determining the user's intentions and needs. Gender classification processing and unified processing are used to identify the user's gender and output gender information, and then unified processing is combined with semantic information to generate personalized responses for the user.

[0050] The personalized response submodule of the response generation module generates personalized response content that meets the user's needs based on the user's personal preferences and historical interaction records. This helps to improve user satisfaction and the interactive experience of the system. The response generation module generates the corresponding response text or voice signal based on the output results of the semantic understanding module and the information of the personalized response submodule.

[0051] The output module converts the response text into a voice signal and plays it to the user through the speaker. This is the last link in the voice interaction system, ensuring that the user can clearly hear the system's response.

[0052] The control module, as the "brain" of the entire system, is responsible for coordinating and controlling the data interaction and function execution between modules. It ensures that each module can work together according to the predetermined process and rules, while extracting and constructing the output acoustic features of the filter and using them for model training to improve the accuracy and robustness of speech recognition.

[0053] Embodiment 2;

[0054] A voice interaction method of a voice interaction system comprises the following steps:

[0055] Step S1: The voice acquisition module collects the user's voice input, converts the collected voice signal into a digital signal, and inputs the digital signal into the voice recognition module for voice recognition processing;

[0056] Step S2: After the voice recognition module processes the received digital signal, it converts the voice signal into text information to facilitate subsequent analysis and understanding;

[0057] Step S3: inputting the text information into the semantic understanding module for semantic analysis and understanding to determine the user's intentions and needs;

[0058] Step S4: Generate a corresponding response text or voice signal through a response generation module according to the user's intention and needs;

[0059] Step S5: Output the response text or voice signal to the user through the output module. The control module is connected with each module to coordinate and control the data interaction and function execution between each module.

[0060] In step S2, the speech recognition module further includes a noise suppression submodule for performing noise suppression processing on the digital signal to improve the accuracy of speech recognition.

[0061] Step S3 also includes using the domain knowledge in the knowledge base submodule to assist the semantic understanding module in performing intent recognition and key information extraction.

[0062] Step S4 also includes using a personalized response submodule to generate personalized response content according to the user's personal preferences and historical interaction records.

[0063] The working principle of the embodiment of the present invention is as follows: the noise suppression submodule performs noise suppression processing on the digital signal, removes interference factors such as background noise and echo, and improves the accuracy of speech recognition. The knowledge base submodule of the semantic understanding module is used to store domain knowledge related to user intentions and key information, providing rich background information for semantic understanding and intention recognition, while the third main control module is responsible for controlling the semantic understanding process, inputting text information into the knowledge base for comparison and analysis, and determining the user's intentions and needs.

[0064] Working principle: The voice input of the user is collected through the voice collection module, and the collected voice signal is converted into a digital signal, and then input into the voice recognition module for voice recognition processing. After the voice recognition module processes the received digital signal, it converts the voice signal into text information for subsequent analysis and understanding. The text information is input into the semantic understanding module for semantic analysis and understanding to determine the user's intentions and needs. According to the user's intentions and needs, the corresponding response text or voice signal is generated through the response generation module. The response text or voice signal is output to the user through the output module. The control module is connected with each module to coordinate and control the data interaction and function execution between each module.

[0065] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and alterations may be made to the embodiments without departing from the principles and spirit thereof, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A voice interaction system, comprising a voice acquisition module, a voice recognition module, a semantic understanding module, a response generation module, an output module and a control module, characterized in that: The voice acquisition module is used to acquire the user's voice input and convert the acquired voice signal into a digital signal; The speech recognition module cooperates with the speech acquisition module to receive digital signals, perform speech recognition processing on the digital signals, and convert the digital signals into text information; The semantic understanding module cooperates with the speech recognition module to receive text information, perform semantic analysis and understanding on the text information, and determine the user's intention and needs; The response generation module is connected to the semantic understanding module to generate corresponding response text or voice signal according to the user's intention and needs; The output module cooperates with the response generation module to output the response text or voice signal to the user; The control module is connected to the other modules and is used to coordinate and control the data interaction and function execution between the modules.

2. The voice interaction method of a voice interaction system according to claim 1, characterized in that: The following steps are involved: Step S1: The voice acquisition module acquires the user's voice input, converts the acquired voice signal into a digital signal, and inputs the digital signal into the voice recognition module for voice recognition processing; Step S2: After the speech recognition module processes the received digital signal, it converts the speech signal into text information to facilitate subsequent analysis and understanding; Step S3: inputting the text information into the semantic understanding module for semantic analysis and understanding to determine the user's intentions and needs; Step S4: generating a corresponding response text or voice signal through the response generation module according to the user's intention and needs; Step S5: outputting the response text or voice signal to the user through the output module. The control module is connected with each module to coordinate and control the data interaction and function execution between each module.

3. A voice interaction system according to claim 1, characterized in that: The voice collection module includes a sound pickup module and a first main control module. The sound pickup module includes a plurality of microphones arranged in an array for receiving sound. The voice recognition module also includes a USB access determination module.

4. A voice interaction system according to claim 1, characterized in that: The speech recognition module includes a CPU control processing module and a second main control module, and adopts a deep learning algorithm to perform speech recognition processing, and can recognize speech signals in multiple languages ​​and accents.

5. A voice interaction system according to claim 1, characterized in that: The semantic understanding module includes a knowledge base submodule and a third main control module, which are used to store domain knowledge related to user intentions and key information, and assist the semantic understanding module in intention recognition and key information extraction. The third main control module is electrically connected to the CPU control processing module; the speech recognition processing also includes gender classification processing and unified processing, which identifies the user's gender and outputs gender information, and uniformly processes gender information and semantic information to obtain personalized response information for the voice information.

6. A voice interaction system according to claim 1, characterized in that: The response generation module includes a personalized response submodule, which is used to generate personalized response content according to the user's personal preferences and historical interaction records; the output module includes an audio port and a speaker, which is used for the final transmission of sound.

7. A voice interaction system according to claim 1, characterized in that: The control module is electrically connected to the CPU control processing module. The control module also includes a filter module for extracting and constructing the output acoustic features of the filter, and inputting the output acoustic features of the filter and the file into the overall model for model training until the model converges.

8. A voice interaction method according to claim 2, characterized in that: In step S2, the speech recognition module further includes a noise suppression submodule for performing noise suppression processing on the digital signal to improve the accuracy of speech recognition.

9. A voice interaction method according to claim 2, characterized in that: The step S3 also includes using the domain knowledge in the knowledge base submodule to assist the semantic understanding module in performing intent recognition and key information extraction.

10. A voice interaction method according to claim 2, characterized in that: The step S4 also includes using a personalized response submodule to generate personalized response content according to the user's personal preferences and historical interaction records.

Images