To-do task generation method and device

Abstract

The invention discloses a to-do task generation method and device, and relates to the technical field of voice recognition, and the method comprises the steps: segmenting a recording audio signal, and carrying out the feature extraction to obtain recording audio segment features; grouping the recorded audio segment features based on the similarity in the similarity matrix, and taking the recorded audio segment features in each recorded audio segment feature group as a cluster; calculating the similarity between the clusters based on the similarity matrix; clustering is carried out based on the similarity between the clusters, and a clustering tree is obtained; distributing speaker labels based on the clustering tree to obtain a target recording audio signal segment, and performing voice-to-text processing on the target recording audio signal segment to obtain text content corresponding to the target recording audio signal segment; according to the keyword table, the historical task template and the text content, dynamically constructing cue words; taking the text generation type as a task type corresponding to a pre-loaded model; and generating a to-do task by using a task generation model according to the cue word. In this way, the efficiency of generating the to-do task is improved.

Description

Technical Field

[0001] The present application relates to the field of speech recognition technology, and in particular to a method and device for generating to-do tasks. Background Art

[0002] With the development of information technology, office automation systems are increasingly being used in enterprises, improving office efficiency.

[0003] Traditional office automation systems include multiple functional modules, each designed to perform different functions. For example, the meeting management module manages and publishes meetings and meeting minutes, while the to-do task management module manages the system's to-do tasks. However, in traditional office automation systems, manual work is required to extract to-do items from meeting minutes in the meeting management module and then publish them within the to-do task management module. However, relying solely on manual work prevents timely updates of meeting minutes and to-do task publication, resulting in low office efficiency. Summary of the Invention

[0004] In view of this, the present application provides a method and device for generating to-do tasks, which are used to solve the problem of low office efficiency in the prior art.

[0005] The purpose of this application can be achieved through the following technical solutions: The first aspect of the present application is to provide a method for generating to-do tasks, comprising: Obtaining a recording file, which includes a recording audio signal; Segmenting the recorded audio signal to obtain multiple recorded audio signal segments; Extract features from the recorded audio signal segment to obtain features of the recorded audio segment; Calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; Grouping the recorded audio segment features based on the similarity in the similarity matrix to obtain multiple recorded audio segment feature groups, and treating the recorded audio segment features in each recorded audio segment feature group as a cluster; Calculate the similarity between clusters based on the similarity matrix; Clustering is performed based on the similarity between clusters to obtain a cluster tree; Assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; Perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; Get keyword lists and historical task templates; Dynamically construct prompt words based on keyword lists, historical task templates, and text content; Use the text generation type as the task type corresponding to the pre-loaded task generation model; Based on the prompt words, the task generation model is used to generate to-do tasks.

[0006] In an optional embodiment, before extracting features from the recorded audio signal segment to obtain the features of the recorded audio segment, the method further includes: Performing a short-time Fourier transform on the recorded audio signal segment to obtain a recorded audio spectrum, where the recorded audio spectrum includes a recorded audio amplitude spectrum and a recorded audio phase spectrum; Perform noise estimation on the recorded audio amplitude spectrum to obtain the noise amplitude spectrum; Performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum; Obtaining a recorded audio enhancement spectrum based on the recorded audio enhancement amplitude spectrum and the recorded audio phase spectrum; Performing an inverse short-time Fourier transform on the enhanced spectrum of the recorded audio to obtain an enhanced signal of the recorded audio; Perform feature extraction on the recorded audio signal segment to obtain the recorded audio segment features, including: Feature extraction is performed on the recorded audio enhancement signal to obtain the recorded audio segment features.

[0007] In an optional embodiment, performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum includes: Based on the noise amplitude spectrum, the following formula is used to perform a spectral subtraction operation on the recorded audio amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum:

[0008] in, is the audio enhancement amplitude spectrum of the recorded audio signal at frequency f, is the amplitude spectrum of the recorded audio signal at frequency f, is the noise amplitude spectrum at frequency f, is the oversubtraction factor, is the protection factor.

[0009] In an optional embodiment, before performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum, the method further includes: Calculate the signal-to-noise ratio of the recorded audio signal at frequency f; Get the noise intensity of the recorded audio signal at frequency f; Adjusting the oversubtraction factor based on the signal-to-noise ratio, the noise intensity, and the initial value of the oversubtraction factor; The protection factor is adjusted based on the signal-to-noise ratio, the noise intensity, and the initial value of the protection factor.

[0010] In an optional embodiment, obtaining the noise intensity of the recorded audio signal at a frequency f includes: Obtaining a midline of the recorded audio signal at a frequency of f, where the midline is determined based on a maximum line of the recorded audio signal at a frequency of f and a minimum line of the recorded audio signal at a frequency of f; The noise intensity of the recorded audio signal at a frequency f is calculated based on the signal value on the center line, the average value of all signal values ​​on the center line, and the number of signal values ​​on the center line.

[0011] In an optional embodiment, before assigning speaker labels based on the clustering tree to obtain the target recorded audio signal segment, the method further includes: Obtain voiceprint samples of participants; Based on the voiceprint samples, a voiceprint sample library is constructed.

[0012] In an optional embodiment, assigning speaker labels based on a clustering tree to obtain a target recorded audio signal segment includes: Based on the voiceprint sample library, a speaker label is assigned to each cluster in the clustering tree.

[0013] In an optional embodiment, the method further includes: Based on the received editing operation, the text content is edited.

[0014] In an optional embodiment, the method further includes: converting the format of the to-do tasks.

[0015] A second aspect of the present application is to provide a to-do task generation device, characterized by comprising: A first acquisition module is used to acquire a recording file, where the recording file includes a recording audio signal; A segmentation module, used to segment the recorded audio signal to obtain multiple recorded audio signal segments; The extraction module is used to extract features from the recorded audio signal segment to obtain the features of the recorded audio segment; The first calculation module is used to calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; a grouping module, configured to group the recorded audio segment features based on the similarity in the similarity matrix to obtain a plurality of recorded audio segment feature groups, and to group the recorded audio segment features in each recorded audio segment feature group as a cluster; A second calculation module is used to calculate the similarity between clusters based on the similarity matrix; Clustering module, used to perform clustering based on the similarity between clusters to obtain a cluster tree; An assignment module, configured to assign speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; The speech-to-text module is used to perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; The second acquisition module is used to obtain the keyword table and historical task template; A construction module is used to dynamically construct prompt words based on the keyword table, historical task templates and text content; A determination module, configured to use a text generation type as a task type corresponding to a pre-loaded task generation model; The generation module is used to generate to-do tasks based on prompt words using the task generation model.

[0016] The third aspect of the present application is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the method of the first aspect when executing the computer program.

[0017] The fourth aspect of the present application is to provide a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method of the first aspect.

[0018] Compared with the existing technology, the to-do task generation method provided by this application segments the recorded audio signal and performs feature extraction to obtain the recorded audio segment features; groups the recorded audio segment features based on the similarity in the calculated similarity matrix to obtain multiple recorded audio segment feature groups, and regards the recorded audio segment features in each recorded audio segment feature group as a cluster; calculates the similarity between clusters based on the similarity matrix; clusters based on the similarity between clusters to obtain a clustering tree; assigns speaker labels based on the clustering tree to obtain the target recorded audio signal segment, and then performs speech-to-text processing on it to obtain the text content corresponding to the target recorded audio signal segment; dynamically constructs prompt words based on the keyword table, historical task templates and text content; uses the text generation type as the task type corresponding to the pre-loaded model; and generates to-do tasks using the task generation model based on the prompt words. In this way, the efficiency of to-do task generation is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0020] Figure 1 A flowchart of a method for generating to-do tasks provided in an embodiment of the present application; Figure 2 Another flowchart of the method for generating to-do tasks provided in an embodiment of the present application; Figure 3 A structural block diagram of a to-do task generation device provided in an embodiment of the present application; Figure 4 A structural block diagram of an electronic device for implementing a method for generating to-do tasks provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] It should be understood that in the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and subsequent associated objects are in an "or" relationship. "Including A, B and / or C" means including any one, any two, or any three of A, B, and C.

[0024] It should be understood that in the embodiments of the present application, "B corresponding to A," "B corresponding to A," "A corresponds to B," or "B corresponds to A" means that B is associated with A and B can be determined based on A. Determining B based on A does not mean determining B based solely on A; B can also be determined based on A and / or other information.

[0025] In order to solve the technical problems existing in the related art, the embodiments of the present application provide a method and device for generating to-do tasks.

[0026] The to-do task generation method provided in the embodiment of the present application can be executed by an electronic device, which can be a terminal or a server. The terminal can be a terminal device such as a smart phone, a tablet computer, or a laptop computer. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides cloud services, cloud databases, cloud computing, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as basic cloud computing services such as big data and artificial intelligence platforms. It can be understood that the present application does not specifically limit the execution entity of the to-do task generation method.

[0027] The technical solution of the present application is described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments described below are used to explain the technical solution of the present application and are not intended to be used as limitations for actual use.

[0028] In order to solve the technical problems existing in the related art, the embodiment of the present application provides a method for generating to-do tasks, such as Figure 1 As shown, Figure 1 A flowchart of a method for generating to-do tasks provided in an embodiment of the present application. It should be noted that the steps shown may be performed in a different logical order than that shown in the flowchart of the method. The method may include the following steps S101 to S113.

[0029] Step S101: Obtain a recording file.

[0030] In an optional embodiment, the recording file may be obtained by recording the voices of conference participants using a recording device during the conference, and the recording file includes a recorded audio signal.

[0031] It should be noted that all participants were informed in advance of the recording and their consent was obtained from all participants that this meeting would be recorded.

[0032] In a specific embodiment, the meeting convener logs into the office automation system, captures voice through a microphone, and pushes the recorded voice to the voiceprint recognition system in real time.

[0033] In a specific embodiment, a communication connection between the office automation system and the voiceprint recognition system is established based on a full-duplex communication protocol.

[0034] In this step, the voices of the participants are recorded during the meeting to generate a recording file so that to-do tasks can be generated directly based on the recording file later.

[0035] In another optional embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: obtaining voiceprint samples of the participants; and constructing a voiceprint sample library based on the voiceprint samples.

[0036] It should be noted that if no speaker is matched in the voiceprint sample library, the current speaker is modified in real time, and the voiceprint of the speaker is saved in the voiceprint sample library.

[0037] In a specific embodiment, voiceprint samples are stored in a voiceprint sample library in the form of sets, and different sets correspond to different meetings.

[0038] In this embodiment, a voiceprint sample library is constructed so that the identities of the participants can be subsequently identified based on the voiceprint sample library.

[0039] Step S102: segmenting the recorded audio signal to obtain a plurality of recorded audio signal segments.

[0040] In an optional embodiment, the recorded audio signal is segmented according to preset time intervals, for example, the preset time intervals are 200 ms (milliseconds), 250 ms, and 300 ms.

[0041] Step S103: extracting features from the recorded audio signal segment to obtain features of the recorded audio segment.

[0042] In an optional embodiment, VAD (Voice Activity Detector) is used to extract features from the recorded audio signal segment to obtain features of the recorded audio segment.

[0043] In an optional embodiment, before extracting features from the recorded audio signal segment to obtain the features of the recorded audio segment, noise processing may be performed on the recorded audio signal segment.

[0044] In a specific embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: Performing a short-time Fourier transform on a recorded audio signal segment to obtain a recorded audio spectrum, which includes a recorded audio amplitude spectrum and a recorded audio phase spectrum; performing noise estimation on the recorded audio amplitude spectrum to obtain a noise amplitude spectrum; performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain a recorded audio enhanced amplitude spectrum; obtaining a recorded audio enhanced spectrum based on the recorded audio enhanced amplitude spectrum and the recorded audio phase spectrum; performing an inverse short-time Fourier transform on the recorded audio enhanced spectrum to obtain a recorded audio enhanced signal; Correspondingly, feature extraction is performed on the recorded audio signal segment to obtain the recorded audio segment feature, which specifically includes the following steps: feature extraction is performed on the recorded audio enhancement signal to obtain the recorded audio segment feature.

[0045] In another specific embodiment, after obtaining the recorded audio spectrum, the non-speech segment in the recorded audio signal segment is identified by VAD, so that noise estimation is subsequently performed based on the non-speech segment.

[0046] In another specific embodiment, MCRA (Minimum Controlled Recursive Averaging) is used to perform noise estimation on the recorded audio amplitude spectrum to obtain the noise amplitude spectrum.

[0047] In a more specific embodiment, performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum includes: Based on the noise amplitude spectrum, the following formula is used to perform a spectral subtraction operation on the recorded audio amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum: (1); in, is the audio enhancement amplitude spectrum of the recorded audio signal at frequency f, is the amplitude spectrum of the recorded audio signal at frequency f, is the noise amplitude spectrum at frequency f, is the oversubtraction factor, is the protection factor.

[0048] In a more specific embodiment, and It is set based on actual needs. For example, when the meeting is held in a conference room, Set to 2.0, Set to 0.01. When the meeting is held at the construction site, Set to 3.0, Set to 0.001. When the meeting is held in KTV, Set to 1.5, Set to 0.05.

[0049] The physical meaning of is to eliminate the frequency band components dominated by noise, The physical meaning of is to ensure that the result is not negative, avoid imaginary spectrum, and retain trace noise. The physical meaning of is to convert the energy domain into the amplitude domain and maintain the waveform continuity when combined with the original phase spectrum.

[0050] In a more specific embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: Calculate the signal-to-noise ratio of the recorded audio signal at a frequency of f; obtain the noise intensity of the recorded audio signal at a frequency of f; adjust the subtraction factor based on the signal-to-noise ratio, the noise intensity, and the initial value of the subtraction factor; and adjust the protection factor based on the signal-to-noise ratio, the noise intensity, and the initial value of the protection factor.

[0051] In a more specific embodiment, obtaining the noise intensity of the recorded audio signal at a frequency of f specifically includes the following steps: obtaining the center line of the recorded audio signal at a frequency of f; and calculating the noise intensity of the recorded audio signal at a frequency of f based on the signal value located on the center line, the average value of all signal values ​​located on the center line, and the number of signal values ​​located on the center line.

[0052] It should be noted that the center line is determined based on the maximum line of the recorded audio signal when the frequency is f and the minimum line of the recorded audio signal when the frequency is f, wherein the maximum line is a line formed by the signal maximum of the recorded audio signal when the frequency is f, and the minimum line is a line formed by the signal minimum of the recorded audio signal when the frequency is f.

[0053] In a more specific embodiment, based on the signal value on the midline, the average of all signal values ​​on the midline, and the number of signal values ​​on the midline, the noise intensity of the recorded audio signal at a frequency of f is calculated using the following formula: (2); Where Q is the noise intensity of the recorded audio signal at frequency f, G is the number of signal values ​​on the midline, is the i-th signal value on the midline, is the average of all signal values ​​lying on the center line.

[0054] In a more specific embodiment, based on the signal-to-noise ratio, the noise intensity, and the initial value of the oversubtraction factor, the oversubtraction factor is adjusted using the following formula to obtain an adjusted oversubtraction factor: (3); (4); in, is the adjusted oversubtraction factor, is the initial value of the oversubtraction factor, and Q is the noise intensity.

[0055] It should be noted that when the signal-to-noise ratio is higher than the preset threshold, the over-subtraction factor is adjusted using formula (4), and when the signal-to-noise ratio is lower than the preset threshold, the over-subtraction factor is adjusted using formula (3).

[0056] In a more specific embodiment, based on the signal-to-noise ratio, the noise intensity, and the initial value of the protection factor, the protection factor is adjusted using the following formula to obtain an adjusted protection factor: (5); (6); in, is the adjusted protection factor, is the initial value of the protection factor, and Q is the noise intensity.

[0057] It should be noted that when the signal-to-noise ratio is higher than the preset threshold, the over-subtraction factor is adjusted using formula (6), and when the signal-to-noise ratio is lower than the preset threshold, the over-subtraction factor is adjusted using formula (5).

[0058] In this embodiment, in the case of low signal-to-noise ratio, the noise is strong and the signal is weak, so a stronger noise estimation is needed to better separate the noise. It can improve the noise suppression ability and thus reduce the impact of noise; Control the retention of the signal, increase It can retain more signal information while suppressing noise. Therefore, when the signal-to-noise ratio is low, increasing This can avoid misidentification of the boundary between noise and signal.

[0059] In the case of high signal-to-noise ratio, the noise is small and the signal is strong, so there is no need to suppress the noise too much. This helps reduce over-suppression of the signal and avoids introducing more distortion. It helps to avoid over-smoothing the details of the signal and keep the original shape of the signal, especially in the case of high signal-to-noise ratio, where the noise has less impact. This ensures better signal retention.

[0060] Step S104: Calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix.

[0061] It should be noted that the features of different audio segments of the same speaker are close in distance in the vector space.

[0062] In an optional embodiment, the cosine similarity between the features of the recorded audio segments is calculated to obtain a similarity matrix.

[0063] Step S105: grouping the recorded audio segment features based on the similarities in the similarity matrix to obtain a plurality of recorded audio segment feature groups, and taking the recorded audio segment features in each recorded audio segment feature group as a cluster.

[0064] In an optional embodiment, based on the similarity in the similarity matrix, a clustering algorithm is used to group the recorded audio segment features to obtain multiple recorded audio segment feature groups, and the recorded audio segment features in each recorded audio segment feature group are regarded as a cluster.

[0065] Step S106: Calculate the similarity between clusters based on the similarity matrix.

[0066] In an optional embodiment, based on the similarity matrix, the maximum similarity is used to calculate the similarity between clusters. In addition, other methods can be used to calculate the similarity between clusters, which is not limited in this application.

[0067] Step S107: clustering is performed based on the similarity between clusters to obtain a cluster tree.

[0068] In an optional embodiment, a hierarchical clustering algorithm is used to perform clustering based on the similarity between clusters to obtain a cluster tree, wherein the cluster tree reflects the acoustic similarity relationship between different speakers.

[0069] Step S108: assigning speaker labels based on the clustering tree to obtain target recorded audio signal segments.

[0070] It should be noted that the target recorded audio signal segment includes the corresponding speaker label.

[0071] In an optional embodiment, assigning speaker labels based on a clustering tree to obtain a target recorded audio signal segment specifically includes the following steps: Based on the voiceprint sample library, a speaker label is assigned to each cluster in the clustering tree.

[0072] That is, clusters are matched with real names, and real names are used as speaker labels.

[0073] Step S109: performing speech-to-text processing on the target recorded audio signal segment to obtain text content corresponding to the target recorded audio signal segment.

[0074] In an optional embodiment, ASR (Automatic Speech Recognition) is used to perform speech-to-text processing on the target recorded audio signal segment to obtain text content corresponding to the target recorded audio signal segment.

[0075] Step S110: Obtain a keyword table and historical task templates.

[0076] In an optional embodiment, the keyword table is maintained through a maintainable interface, and the keyword table includes a sequence number, a keyword, and a number of times the keyword is used.

[0077] In an optional embodiment, keywords in the keyword table can be added, modified, and deleted later, and keywords can also be imported in batches through templates.

[0078] In an optional embodiment, the data source of the historical task template is the to-do tasks generated historically, and the target quantity is controlled according to the utilization rate and capacity. The template content includes the time node, responsible person and task content. For example, responsible person A completes task B before the deadline.

[0079] Step S111: dynamically construct prompt words according to the keyword table, historical task templates and text content.

[0080] The prompt word refers to the text entered when interacting with the task generation model, which is used to guide the task generation model to generate specific outputs. Among them, the task generation model refers to a "large parameter" model trained using large-scale data and powerful computing power. It has a high degree of versatility and generalization ability and can be applied to fields such as natural language processing, image recognition, and speech recognition. It can be divided into large language models, large visual models, multimodal large models, and basic large models. For example, the task generation large model in the embodiment of the present application can be a large language model.

[0081] In an optional embodiment, the task and output format of the prompt word can be determined according to the meeting minutes format of the target meeting and the target text content, so that the prompt word can be constructed.

[0082] Exemplarily, the meeting minutes format of the target meeting is selected as the output format of the prompt word, and the prompt word is constructed based on the target text content and the meeting minutes format.

[0083] Step S112: The text generation type is used as the task type corresponding to the pre-loaded task generation model.

[0084] In an optional embodiment, a task generation model is pre-loaded, and the training goal is to "generate structured tasks from unstructured text." The task type of the model can be conditional text generation or information extraction.

[0085] Step S113: Generate to-do tasks using a task generation model based on the prompt word.

[0086] In an optional embodiment, the prompt word is input into the task generation model, reasoning is performed, and a structured to-do task list is output.

[0087] In another optional embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: The text content is preprocessed, where the preprocessing operation includes at least one of the following operations: a cleaning operation, a format unification operation, and a sentence segmentation operation.

[0088] In a specific embodiment, the cleaning operation may involve removing irrelevant content such as greetings and small talk from the text. The formatting operation refers to standardizing the date / time format of the text. The sentence segmentation operation involves breaking long paragraphs into short, semantically independent sentences, using a mixed sentence segmentation process (i.e., using punctuation and semantics).

[0089] In another optional embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: Based on the received editing operation, the text content is edited.

[0090] In this embodiment, the text content is modified to improve the accuracy of subsequent task generation.

[0091] In another optional embodiment, the method for generating to-do tasks provided in the embodiment of the present application further includes the following steps: converting the format of the to-do tasks so that the generated tasks can be "machine-understood" and "system-used".

[0092] like Figure 2 As shown, the embodiment of the present application also provides a method for generating to-do tasks, which specifically includes the following steps: Step 1: Get the keyword table set by calling the keyword API (Application Programming Interface).

[0093] Step 2: Get the historical task template collection through the task template API.

[0094] Step 3: Enter the target text content.

[0095] Step 4: Dynamically construct prompt words based on the keyword table, historical task templates and text content.

[0096] Through the above design, the system can respond to changes in the keyword table and historical task templates in real time, keep the generated results consistent with the latest enterprise standards, and implement modular prompt word component management; Step 5: Load the task generation model and select the text generation type; Step 6: Use the task generation model to extract the task to-do list based on the dynamically constructed prompt words. This process generates 1 to N different to-do tasks. Step 7: Generate the corresponding format of the to-do task list through the rule engine.

[0097] It should be noted that different rule engines can implement different format conversions.

[0098] Corresponding to the method for generating to-do tasks provided in the embodiment of the present application, the embodiment of the present application also provides a device for generating to-do tasks, such as Figure 3 As shown, the to-do task generating device includes: The first acquisition module 301 is used to acquire a recording file, which includes a recording audio signal; A segmentation module 302 is used to segment the recorded audio signal to obtain multiple recorded audio signal segments; Extraction module 303, used to extract features from the recorded audio signal segment to obtain features of the recorded audio segment; A first calculation module 304 is used to calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; A grouping module 305 is configured to group the recorded audio segment features based on the similarities in the similarity matrix to obtain a plurality of recorded audio segment feature groups, and to group the recorded audio segment features in each recorded audio segment feature group as a cluster; A second calculation module 306 is used to calculate the similarity between clusters based on the similarity matrix; Clustering module 307, used to perform clustering based on the similarity between clusters to obtain a cluster tree; an assignment module 308 for assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; The speech-to-text module 309 is configured to perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; The second acquisition module 310 is used to acquire a keyword table and a historical task template; A construction module 311 is used to dynamically construct prompt words based on the keyword table, the historical task template and the text content; A determination module 312 is configured to use the text generation type as a task type corresponding to a pre-loaded task generation model; The generation module 313 is used to generate to-do tasks according to the prompt words using the task generation model.

[0099] Corresponding to the method for generating to-do tasks provided in the embodiment of the present application, the embodiment of the present application also provides an electronic device for executing the method for generating to-do tasks, such as Figure 4 As shown, the electronic device includes: a processor 401; and a memory 402, which is used to store a program of a to-do task generation method. After the device is powered on and the program of the to-do task generation method is run by the processor, the following steps are performed: Obtaining a recording file, which includes a recording audio signal; Segmenting the recorded audio signal to obtain multiple recorded audio signal segments; Extract features from the recorded audio signal segment to obtain features of the recorded audio segment; Calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; Grouping the recorded audio segment features based on the similarity in the similarity matrix to obtain multiple recorded audio segment feature groups, and treating the recorded audio segment features in each recorded audio segment feature group as a cluster; Calculate the similarity between clusters based on the similarity matrix; Clustering is performed based on the similarity between clusters to obtain a cluster tree; Assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; Perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; Get keyword lists and historical task templates; Dynamically construct prompt words based on keyword lists, historical task templates, and text content; Use the text generation type as the task type corresponding to the pre-loaded task generation model; Based on the prompt words, the task generation model is used to generate to-do tasks.

[0100] Corresponding to the method for generating to-do tasks provided in the embodiment of the present application, the embodiment of the present application further provides a computer-readable storage medium storing a program for the method for generating to-do tasks, which is executed by a processor to perform the following steps: Obtaining a recording file, which includes a recording audio signal; Segmenting the recorded audio signal to obtain multiple recorded audio signal segments; Extract features from the recorded audio signal segment to obtain features of the recorded audio segment; Calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; Grouping the recorded audio segment features based on the similarity in the similarity matrix to obtain multiple recorded audio segment feature groups, and treating the recorded audio segment features in each recorded audio segment feature group as a cluster; Calculate the similarity between clusters based on the similarity matrix; Clustering is performed based on the similarity between clusters to obtain a cluster tree; Assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; Perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; Get keyword lists and historical task templates; Dynamically construct prompt words based on keyword lists, historical task templates, and text content; Use the text generation type as the task type corresponding to the pre-loaded task generation model; Based on the prompt words, the task generation model is used to generate to-do tasks.

[0101] Corresponding to the method for generating to-do tasks provided in the embodiment of the present application, the embodiment of the present application further provides a computer program comprising instructions. When the program is executed by a computer, the instructions cause the computer to perform the following steps: Obtaining a recording file, which includes a recording audio signal; Segmenting the recorded audio signal to obtain multiple recorded audio signal segments; Extract features from the recorded audio signal segment to obtain features of the recorded audio segment; Calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; Grouping the recorded audio segment features based on the similarity in the similarity matrix to obtain multiple recorded audio segment feature groups, and treating the recorded audio segment features in each recorded audio segment feature group as a cluster; Calculate the similarity between clusters based on the similarity matrix; Clustering is performed based on the similarity between clusters to obtain a cluster tree; Assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; Perform speech-to-text processing on the target recorded audio signal segment to obtain the text content corresponding to the target recorded audio signal segment; Get keyword lists and historical task templates; Dynamically construct prompt words based on keyword lists, historical task templates, and text content; Use the text generation type as the task type corresponding to the pre-loaded task generation model; Based on the prompt words, the task generation model is used to generate to-do tasks.

[0102] It should be noted that for the detailed description of the to-do task generation device, electronic device and computer-readable storage medium provided in the embodiment of the present application, reference can be made to the relevant description of the to-do task generation method embodiment provided in the embodiment of the present application, and no further details will be given here.

[0103] Although the present application is disclosed as above with the preferred embodiments, it is not intended to limit the present application. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

[0104] In a typical configuration, an electronic device includes one or more processors (Central Processing Unit), input / output interfaces, network interfaces, and memory.

[0105] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash memory. Memory is an example of a computer-readable medium.

[0106] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can be implemented using any method or technology to store information. The information can be computer-readable operations, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, programmable analog modules (PAMs), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RANM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPRM), flash memory or other memory technologies, compact disc read-only memory (CDROM), digital versatile disc (DVCD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0107] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, compact disc read-only memory, optical storage, etc.) containing computer-usable program code.

[0108] Although the present application is disclosed as above with the preferred embodiments, it is not intended to limit the present application. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined in the present application.

Claims

1. A method for generating to-do tasks, characterized in that: include: Obtaining a recording file, wherein the recording file includes a recording audio signal; Segmenting the recorded audio signal to obtain a plurality of recorded audio signal segments; Extracting features from the audio recording signal to obtain audio recording features. Calculating the similarity between the features of the recorded audio segments to obtain a similarity matrix; Grouping the recorded audio segment features based on the similarities in the similarity matrix to obtain a plurality of recorded audio segment feature groups, and treating the recorded audio segment features in each of the recorded audio segment feature groups as a cluster; Calculating the similarity between the clusters based on the similarity matrix; Perform clustering based on the similarity between the clusters to obtain a cluster tree; assigning speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; Performing speech-to-text processing on the target recorded audio signal segment to obtain text content corresponding to the target recorded audio signal segment; Get keyword lists and historical task templates; Dynamically construct prompt words according to the keyword table, the historical task template and the text content; Use the text generation type as the task type corresponding to the pre-loaded task generation model; According to the prompt word, the task generation model is used to generate to-do tasks.

2. The method for generating pending tasks according to claim 1, wherein: Before extracting features from the recorded audio signal segment to obtain features of the recorded audio segment, the method further includes: Performing a short-time Fourier transform on the recorded audio signal segment to obtain a recorded audio spectrum, wherein the recorded audio spectrum includes a recorded audio amplitude spectrum and a recorded audio phase spectrum; Performing noise estimation on the recorded audio amplitude spectrum to obtain a noise amplitude spectrum; Performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain an enhanced amplitude spectrum of the recorded audio; Obtaining a recorded audio enhancement spectrum based on the recorded audio enhancement amplitude spectrum and the recorded audio phase spectrum; Performing an inverse short-time Fourier transform on the recorded audio enhancement spectrum to obtain a recorded audio enhancement signal; Extracting features from the recorded audio signal segment to obtain features of the recorded audio segment includes: Feature extraction is performed on the recorded audio enhancement signal to obtain recorded audio segment features.

3. The method for generating pending tasks according to claim 2, wherein: The performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum includes: Based on the noise amplitude spectrum, a spectrum subtraction operation is performed on the recorded audio amplitude spectrum using the following formula to obtain the recorded audio enhanced amplitude spectrum: in, is the recorded audio enhanced amplitude spectrum of the recorded audio signal at a frequency of f, is the recorded audio amplitude spectrum of the recorded audio signal at a frequency of f, is the noise amplitude spectrum of the noise at frequency f, is the oversubtraction factor, is the protection factor.

4. The method for generating pending tasks according to claim 3, wherein: Before performing a spectral subtraction operation on the recorded audio amplitude spectrum based on the noise amplitude spectrum to obtain the recorded audio enhanced amplitude spectrum, the method further includes: Calculating the signal-to-noise ratio of the recorded audio signal at a frequency of f; Obtaining the noise intensity of the recorded audio signal at a frequency of f; adjusting the subtraction factor based on the signal-to-noise ratio, the noise intensity, and an initial value of the subtraction factor; The protection factor is adjusted based on the signal-to-noise ratio, the noise intensity, and an initial value of the protection factor.

5. The method for generating pending tasks according to claim 4, wherein: The obtaining of the noise intensity of the recorded audio signal at a frequency f includes: Obtaining a midline of the recorded audio signal at a frequency of f, wherein the midline is determined based on a maximum line of the recorded audio signal at a frequency of f and a minimum line of the recorded audio signal at a frequency of f; The noise intensity of the recorded audio signal at a frequency of f is calculated based on the signal value located on the center line, the average value of all signal values ​​located on the center line, and the number of signal values ​​located on the center line.

6. The method for generating pending tasks according to claim 1, wherein: Before assigning speaker labels based on the clustering tree to obtain target recorded audio signal segments, the method further includes: Obtain voiceprint samples of participants; Based on the voiceprint samples, a voiceprint sample library is constructed.

7. The method for generating pending tasks according to claim 6, wherein: The assigning of speaker labels based on the clustering tree to obtain a target recorded audio signal segment includes: Based on the voiceprint sample library, a speaker label is assigned to each cluster in the clustering tree.

8. The method for generating pending tasks according to claim 1, wherein: Also includes: Based on the received editing operation, the text content is edited.

9. The method for generating pending tasks according to claim 1, wherein: Also includes: Perform format conversion on the to-do task.

10. A device for generating pending tasks, characterized in that: include: A first acquisition module is used to acquire a recording file, wherein the recording file includes a recording audio signal; A segmentation module, configured to segment the recorded audio signal to obtain a plurality of recorded audio signal segments; An extraction module, configured to extract features from the recorded audio signal segment to obtain features of the recorded audio segment; A first calculation module is used to calculate the similarity between the features of the recorded audio segments to obtain a similarity matrix; a grouping module, configured to group the recorded audio segment features based on the similarities in the similarity matrix to obtain a plurality of recorded audio segment feature groups, and to group the recorded audio segment features in each of the recorded audio segment feature groups as a cluster; A second calculation module, configured to calculate the similarity between the clusters based on the similarity matrix; A clustering module, configured to perform clustering based on the similarity between the clusters to obtain a cluster tree; an allocating module, configured to allocate speaker labels based on the clustering tree to obtain a target recorded audio signal segment, wherein the target recorded audio signal segment includes a corresponding speaker label; A speech-to-text module is configured to perform speech-to-text processing on the target recorded audio signal segment to obtain text content corresponding to the target recorded audio signal segment; The second acquisition module is used to obtain the keyword table and historical task template; A construction module, configured to dynamically construct prompt words based on the keyword table, the historical task template, and the text content; A determination module, configured to use a text generation type as a task type corresponding to a pre-loaded task generation model; A generation module is used to generate to-do tasks based on the prompt word using the task generation model.

Images