System function page intelligent scheduling method, system, equipment and medium based on voice recognition

Abstract

The invention discloses a system function page intelligent scheduling method, system and equipment based on voice recognition and a medium, belongs to the technical field of man-machine interaction, and aims to solve the technical problem of how to accurately recognize a voice instruction of a user, intelligently schedule a corresponding system function page and improve operation convenience and user experience. According to the technical scheme, the method comprises the steps of voice collection: collecting a voice instruction of a user, and obtaining a voice signal; voice signal preprocessing: carrying out preprocessing operation of noise reduction and enhancement on the collected voice signals; speech recognition: adopting a recurrent neural network to construct a deep learning model, performing interface processing on speech signals after training, converting acoustic features into a text sequence, performing grammar and punctuation correction on the text sequence, and outputting recognized text data; natural language understanding: analyzing semantic content in the text data, and extracting a user intention by adopting a classification method; intention mapping; and page scheduling.

Description

Technical Field

[0001] The present invention relates to the field of human-computer interaction technology, and in particular to a method, system, device and medium for intelligently scheduling system function pages based on voice recognition. Background Art

[0002] At present, the operation of smart devices mostly relies on touch interfaces, which is not convenient enough when the user's hands are busy or in certain environments. Although there are some voice-based control systems on the market, they can usually only process simple commands and have limitations in scheduling complex function pages, which cannot meet users' needs for efficient and personalized operation of smart devices.

[0003] Therefore, how to accurately identify the user's voice commands and intelligently schedule the corresponding system function pages to improve the convenience of operation and user experience is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The technical task of the present invention is to provide a method, system, device and medium for intelligent scheduling of system function pages based on voice recognition to solve the problem of how to accurately recognize the user's voice commands and intelligently schedule the corresponding system function pages to improve the convenience of operation and user experience.

[0005] The technical task of the present invention is achieved in the following way: a method for intelligently scheduling system function pages based on voice recognition, which uses voice recognition technology to accurately capture and analyze user voice commands, and then intelligently schedules corresponding system function pages; the details are as follows:

[0006] Voice collection: collect user's voice commands and obtain voice signals;

[0007] Speech signal preprocessing: perform noise reduction and enhancement preprocessing operations on the collected speech signals;

[0008] Speech recognition: A recurrent neural network (RNN) is used to build a deep learning model. After training, the speech signal is interfaced and the acoustic features are converted into a text sequence. The text sequence is then corrected for grammar and punctuation, and the recognized text data is output;

[0009] Natural language understanding: parse the semantic content in text data and extract user intent using classification methods;

[0010] Intent mapping: Create a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intent to a specific system function page;

[0011] Page scheduling: Schedule and open the corresponding system function page according to the voice mapping results.

[0012] Preferably, the voice instructions are collected by hardware devices, and the hardware devices include a microphone.

[0013] Preferably, the noise reduction adopts the Wiener filtering method, specifically: different frequency bands are suppressed to different degrees through a linear filter to achieve the purpose of noise reduction.

[0014] Preferably, the enhancement is performed by pre-emphasis processing, specifically: the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and acoustic features are extracted from the speech signal;

[0015] Among them, the acoustic features include Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC);

[0016] The Mel cepstral coefficient extraction process is as follows:

[0017] Framing;

[0018] Periodogram estimation of power spectrum;

[0019] Mel filter bank application;

[0020] Log filter bank energy;

[0021] Discrete Cosine Transform (DCT).

[0022] Preferably, the semantic content in the parsed text data is as follows:

[0023] Perform preprocessing of word segmentation, tagging, grammatical analysis and entity recognition on the text data (excluding non-language elements such as punctuation marks, numbers, special characters in the text, and converting them to lowercase) to obtain the preprocessed text data;

[0024] Perform dependency analysis on the preprocessed text data.

[0025] A system function page intelligent scheduling system based on voice recognition, the system comprising:

[0026] Voice collection module, used to collect user's voice commands and obtain voice signals;

[0027] A speech signal preprocessing module is used to perform noise reduction and enhancement preprocessing operations on the collected speech signal;

[0028] The speech recognition module is used to build a deep learning model using a recurrent neural network (RNN). After training, it interfaces the speech signal and converts the acoustic features into a text sequence. It then corrects the grammar and punctuation of the text sequence and outputs the recognized text data.

[0029] The natural language understanding module is used to parse the semantic content in text data and extract user intent using classification methods;

[0030] The intent mapping module is used to establish a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intent to a specific system function page;

[0031] The page scheduling module is used to schedule and open the corresponding system function page according to the voice mapping results.

[0032] Preferably, the speech signal preprocessing module comprises:

[0033] The noise reduction submodule is used to perform noise reduction on the speech signal. The noise reduction adopts the Wiener filtering method, which is to suppress different frequency bands to different degrees through a linear filter to achieve the purpose of noise reduction.

[0034] The enhancement submodule is used to enhance the speech signal. The enhancement is performed by pre-emphasis. Specifically, the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and acoustic features are extracted from the speech signal;

[0035] Among them, the acoustic features include Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC);

[0036] The Mel cepstral coefficient extraction process is as follows:

[0037] ① Frame division;

[0038] ②Estimation of the periodogram of the power spectrum;

[0039] ③Application of Mel filter bank;

[0040] ④Logarithmic filter bank energy;

[0041] ⑤Discrete cosine transform (DCT).

[0042] Preferably, the natural language understanding module includes:

[0043] The text preprocessing submodule is used to perform word segmentation, tagging, grammatical analysis and entity recognition preprocessing on the text data (excluding non-language elements such as punctuation marks, numbers, special characters in the text, and converting them to lowercase) to obtain the preprocessed text data;

[0044] The dependency analysis submodule is used to perform dependency analysis on the preprocessed text data.

[0045] An electronic device comprising: a memory and at least one processor;

[0046] Wherein, the memory stores a computer program;

[0047] The at least one processor executes the computer program stored in the memory, so that the at least one processor executes the above-mentioned system function page intelligent scheduling method based on voice recognition.

[0048] A computer-readable storage medium stores a computer program, which can be executed by a processor to implement the above-mentioned system function page intelligent scheduling method based on voice recognition.

[0049] The system function page intelligent scheduling method, system, device and medium based on voice recognition of the present invention have the following advantages:

[0050] (i) The present invention uses deep learning algorithms to improve the accuracy of speech recognition, understands the user's natural language instructions through natural language processing technology, can accurately recognize the user's voice instructions, and intelligently schedule the corresponding system function pages to improve the convenience of operation and user experience;

[0051] (ii) The present invention adopts an end-to-end deep learning algorithm to improve the accuracy and response speed of speech recognition;

[0052] (iii) The present invention utilizes natural language processing technology to understand and parse complex instructions;

[0053] (iv) The present invention designs an intelligent context understanding mechanism that can intelligently predict user intentions based on the user's historical operations and current environment;

[0054] (V) The present invention implements multi-round dialogue management and supports continuous interaction of users in complex tasks;

[0055] (6) The present invention achieves high-accuracy recognition of complex voice commands, improving the intelligence of the system and user experience;

[0056] (VII) The present invention uses natural language understanding, and users can interact with the device using natural language without having to memorize specific commands;

[0057] (VIII) The intelligent context understanding mechanism of the present invention can intelligently predict the user's intention based on the user's historical operations and current environment, and can provide personalized services, thereby improving user satisfaction;

[0058] (IX) The present invention manages multiple rounds of dialogues and supports continuous interaction of users in complex tasks, making it smoother for users to perform complex tasks and reducing the number of operation steps. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention is further described below in conjunction with the accompanying drawings.

[0060] Attached Figure 1 The flowchart of the intelligent scheduling method of system function pages based on speech recognition is shown in FIG. DETAILED DESCRIPTION

[0061] The method, system, device and medium for intelligently scheduling system function pages based on speech recognition of the present invention are described in detail below with reference to the accompanying drawings and specific embodiments of the specification.

[0062] Embodiment 1:

[0063] As attached Figure 1 As shown, this embodiment provides a method for intelligently scheduling system function pages based on voice recognition, which uses voice recognition technology to accurately capture and analyze user voice commands, and then intelligently schedules corresponding system function pages; the details are as follows:

[0064] S1. Voice collection: collect user's voice commands and obtain voice signals;

[0065] S2, speech signal preprocessing: performing noise reduction and enhancement preprocessing operations on the collected speech signal;

[0066] S3, speech recognition: A deep learning model is constructed using a recurrent neural network (RNN). After training, the speech signal is interfaced and the acoustic features are converted into a text sequence. The text sequence is then corrected for grammar and punctuation, and the recognized text data is output;

[0067] S4, Natural Language Understanding: Parse the semantic content in text data and extract user intent using classification methods;

[0068] S5, Intent mapping: Establish a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intention to a specific system function page;

[0069] S6. Page scheduling: According to the voice mapping results, schedule and open the corresponding system function page.

[0070] The voice command in step S1 of this embodiment is collected by hardware equipment, and the hardware equipment includes a microphone.

[0071] The noise reduction in step S2 of this embodiment adopts the Wiener filtering method, which is specifically: suppressing different frequency bands to different degrees through a linear filter to achieve the purpose of noise reduction.

[0072] The enhancement in step S2 of this embodiment is performed by pre-emphasis processing, specifically: the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and the acoustic features are extracted from the speech signal;

[0073] Among them, the acoustic features include Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC);

[0074] The Mel cepstral coefficient extraction process is as follows:

[0075] ① Frame division;

[0076] ② Estimation of the periodogram of the power spectrum;

[0077] ③Mel filter bank application;

[0078] ④Logarithmic filter bank energy;

[0079] ⑤Discrete cosine transform (DCT).

[0080] The semantic content of the parsed text data in step S4 of this embodiment is specifically as follows:

[0081] S401, preprocessing the text data by word segmentation, tagging, grammatical analysis and entity recognition (excluding non-language elements such as punctuation marks, numbers, special characters in the text, and converting them to lowercase), and obtaining the preprocessed text data;

[0082] S402: Perform dependency analysis on the preprocessed text data.

[0083] Embodiment 2:

[0084] This embodiment provides a system function page intelligent scheduling system based on voice recognition, the system comprising:

[0085] Voice collection module, used to collect user's voice commands and obtain voice signals;

[0086] A speech signal preprocessing module is used to perform noise reduction and enhancement preprocessing operations on the collected speech signal;

[0087] The speech recognition module is used to build a deep learning model using a recurrent neural network (RNN). After training, it interfaces the speech signal and converts the acoustic features into a text sequence. It then corrects the grammar and punctuation of the text sequence and outputs the recognized text data.

[0088] The natural language understanding module is used to parse the semantic content in text data and extract user intent using classification methods;

[0089] The intent mapping module is used to establish a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intent to a specific system function page;

[0090] The page scheduling module is used to schedule and open the corresponding system function page according to the voice mapping results.

[0091] The speech signal preprocessing module in this embodiment includes:

[0092] The noise reduction submodule is used to perform noise reduction on the speech signal. The noise reduction adopts the Wiener filtering method, which is to suppress different frequency bands to different degrees through a linear filter to achieve the purpose of noise reduction.

[0093] The enhancement submodule is used to enhance the speech signal. The enhancement is performed by pre-emphasis. Specifically, the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and acoustic features are extracted from the speech signal;

[0094] Among them, the acoustic features include Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC);

[0095] The Mel cepstral coefficient extraction process is as follows:

[0096] ① Frame division;

[0097] ②Estimation of the periodogram of the power spectrum;

[0098] ③Application of Mel filter bank;

[0099] ④Logarithmic filter bank energy;

[0100] ⑤Discrete cosine transform (DCT).

[0101] The natural language understanding module in this embodiment includes:

[0102] The text preprocessing submodule is used to perform word segmentation, tagging, grammatical analysis and entity recognition preprocessing on the text data (excluding non-language elements such as punctuation marks, numbers, special characters in the text, and converting them to lowercase) to obtain the preprocessed text data;

[0103] The dependency analysis submodule is used to perform dependency analysis on the preprocessed text data.

[0104] Embodiment 3:

[0105] This embodiment also provides an electronic device, including: a memory and a processor;

[0106] Wherein, the memory stores computer-executable instructions;

[0107] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the system function page intelligent scheduling method based on voice recognition in any embodiment of the present invention.

[0108] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor may be a microprocessor or any conventional processor, etc.

[0109] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the terminal, etc. In addition, the memory can also include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, at least one disk storage period, a flash memory device, or other volatile solid-state storage devices.

[0110] Embodiment 4:

[0111] This embodiment also provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions are loaded by a processor, so that the processor executes the system function page intelligent scheduling method based on voice recognition in any embodiment of the present invention. Specifically, a system or device equipped with a storage medium can be provided, on which a software program code that implements the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0112] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0113] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RYM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer via a communication network.

[0114] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0115] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent scheduling of system function pages based on voice recognition, characterized in that: This method uses voice recognition technology to accurately capture and analyze user voice commands, and then intelligently schedules the corresponding system function pages; the details are as follows: Voice collection: collect user's voice commands and obtain voice signals; Speech signal preprocessing: perform noise reduction and enhancement preprocessing operations on the collected speech signals; Speech recognition: A deep learning model is constructed using a recurrent neural network. After training, the speech signal is interfaced and the acoustic features are converted into a text sequence. The text sequence is then corrected for grammar and punctuation, and the recognized text data is output; Natural language understanding: parse the semantic content in text data and extract user intent using classification methods; Intent mapping: Create a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intent to a specific system function page; Page scheduling: Schedule and open the corresponding system function page according to the voice mapping results.

2. The method for intelligently scheduling system function pages based on voice recognition according to claim 1, characterized in that: Voice commands are collected using hardware devices, including microphones.

3. The system function page intelligent scheduling method based on voice recognition according to claim 1 is characterized in that: The noise reduction adopts the Wiener filtering method, which is: different frequency bands are suppressed to different degrees through a linear filter to achieve the purpose of noise reduction.

4. The system function page intelligent scheduling method based on voice recognition according to claim 1 is characterized in that: The enhancement is performed by pre-emphasis processing, specifically: the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and the acoustic features are extracted from the speech signal; Among them, the acoustic features include Mel cepstral coefficients and linear prediction cepstral coefficients; The Mel cepstral coefficient extraction process is as follows: Framing; Periodogram estimation of power spectrum; Mel filter bank application; Log filter bank energy; Discrete cosine transform.

5. The system function page intelligent scheduling method based on voice recognition according to any one of claims 1 to 4, characterized in that: The semantic content of the parsed text data is as follows: Preprocess the text data by word segmentation, tagging, grammatical analysis and entity recognition to obtain the preprocessed text data; Perform dependency analysis on the preprocessed text data.

6. A system function page intelligent scheduling system based on voice recognition, characterized in that: The system includes: Voice collection module, used to collect user's voice commands and obtain voice signals; A speech signal preprocessing module is used to perform noise reduction and enhancement preprocessing operations on the collected speech signal; The speech recognition module is used to build a deep learning model using a recurrent neural network. After training, it interfaces the speech signal and converts the acoustic features into a text sequence. It then corrects the grammar and punctuation of the text sequence and outputs the recognized text data. The natural language understanding module is used to parse the semantic content in text data and extract user intent using classification methods; The intent mapping module is used to establish a keyword or phrase library for the voice-controlled system, analyze and capture the voice content, and map the user's intent to a specific system function page; The page scheduling module is used to schedule and open the corresponding system function page according to the voice mapping results.

7. The system function page intelligent scheduling system based on voice recognition according to claim 6 is characterized in that: The speech signal preprocessing module includes: The noise reduction submodule is used to perform noise reduction on the speech signal. The noise reduction adopts the Wiener filtering method, which is to suppress different frequency bands to different degrees through a linear filter to achieve the purpose of noise reduction. The enhancement submodule is used to enhance the speech signal. The enhancement is performed by pre-emphasis. Specifically, the first order is implemented by using a FIR high-pass filter, and the pre-emphasis coefficient is between 0.9 and 1.0; and acoustic features are extracted from the speech signal; Among them, the acoustic features include Mel cepstral coefficients and linear prediction cepstral coefficients; The Mel cepstral coefficient extraction process is as follows: ① Frame division; ② Estimation of the periodogram of the power spectrum; ③Application of Mel filter bank; ④Logarithmic filter bank energy; ⑤Discrete cosine transform.

8. The system function page intelligent scheduling system based on voice recognition according to claim 6 or 7, characterized in that: The natural language understanding module includes: The text preprocessing submodule is used to preprocess the text data by word segmentation, tagging, grammatical analysis and entity recognition to obtain the preprocessed text data; The dependency analysis submodule is used to perform dependency analysis on the preprocessed text data.

9. An electronic device, characterized in that: include: memory and at least one processor; Wherein, the memory stores a computer program; The at least one processor executes the computer program stored in the memory, so that the at least one processor executes the system function page intelligent scheduling method based on voice recognition as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which can be executed by a processor to implement the system function page intelligent scheduling method based on voice recognition as described in any one of claims 1 to 5.

Images