A voice processing method, device, equipment and storage medium

Abstract

The present application discloses a speech processing method, apparatus, device and storage medium, wherein the method comprises: obtaining speech to be processed; identifying the speech to be processed using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed; and if the speech to be processed is played, playing the enhanced speech segment in a manner different from playing the regular speech segment.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of speech processing, and relate to but are not limited to a speech processing method, apparatus, device, and storage medium. Background Art

[0002] When a user is chatting on a chat software, he or she receives a long voice message from the other party, which is played repeatedly. The chat software has problems with inaccurate recognition and inability to distinguish key points, resulting in a poor user experience. Summary of the Invention

[0003] In view of this, embodiments of the present application provide a speech processing method, apparatus, device, and storage medium.

[0004] The technical solution of the embodiment of the present application is implemented as follows:

[0005] In a first aspect, an embodiment of the present application provides a speech processing method, the method comprising: obtaining a speech to be processed; identifying the speech to be processed using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed; if the speech to be processed is played, the enhanced speech segment is played in a manner different from that of playing the regular speech segment.

[0006] In a second aspect, an embodiment of the present application provides a speech processing device, comprising: an acquisition module for acquiring speech to be processed; a recognition module for recognizing the speech to be processed using a semantic understanding model, and obtaining at least one regular speech segment and at least one enhanced speech segment in the speech to be processed; and a playback module for playing the speech to be processed, and playing the enhanced speech segment in a manner different from that of playing the regular speech segment.

[0007] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the above method when executing the program.

[0008] In a fourth aspect, an embodiment of the present application provides a storage medium storing executable instructions for causing a processor to execute and implement the above method.

[0009] In the embodiment of the present application, the speech to be processed is first acquired; then, the speech to be processed is recognized using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment from the speech to be processed; finally, when the speech to be processed is played, the enhanced speech segment is played in a different manner from the regular speech segment. In this way, by playing the regular speech segment and the enhanced speech segment in different manners, the user can quickly grasp the key content in a large text paragraph or the key speech in a speech segment. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram of an implementation flow of a speech processing method provided in an embodiment of the present application;

[0011] Figure 2A A schematic diagram of an implementation flow of a speech processing method provided in an embodiment of the present application;

[0012] Figure 2B A schematic diagram of playing a voice to be processed provided in an embodiment of the present application;

[0013] Figure 2C A schematic diagram of playing an enhanced voice segment provided in an embodiment of the present application;

[0014] Figure 3 A schematic diagram of an implementation flow of a method for determining key speech provided in an embodiment of the present application;

[0015] Figure 4 A schematic diagram of an implementation flow of a speech processing method provided in an embodiment of the present application;

[0016] Figure 5 A schematic diagram of the structure of a speech processing device provided in an embodiment of the present application;

[0017] Figure 6 A schematic diagram of a hardware entity of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] To make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the specific technical solutions of the embodiments of the present application will be further described in detail below in conjunction with the drawings in the embodiments of the present application. The following embodiments are used to illustrate the present application but are not intended to limit the scope of the present application.

[0019] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0020] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0022] The present application provides a method for voice processing. Figure 1 As shown, the method includes:

[0023] Step S110: Acquire the speech to be processed;

[0024] Here, the voice to be processed may be the voice acquired by the user using voice receiving software.

[0025] During implementation, the duration of the user-input voice can be determined in units of time. In some embodiments, a voice duration threshold can be set. First, the voice duration is determined based on the acquired voice, and then the voice duration is compared with the voice duration threshold. If the voice duration exceeds the voice duration threshold, the voice is determined as a voice to be processed. In this way, voices with a duration greater than the voice duration threshold can be processed according to user needs, while voices with a duration less than the voice duration threshold can be left unprocessed.

[0026] Step S120: using a semantic understanding model to recognize the speech to be processed, and obtaining at least one regular speech segment and at least one enhanced speech segment in the speech to be processed;

[0027] Here, the regular speech segments may include prepositions, interjections, and pronouns in speech, and the enhanced speech segments may include place nouns, time nouns, and person nouns.

[0028] In some embodiments, the semantic understanding model may be a model for recognizing speech, that is, the semantic understanding model may obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed by recognizing the speech to be processed.

[0029] In some embodiments, the semantic understanding model can be a model for recognizing text, that is, the speech to be processed is first recognized as text to be processed, and then the semantic understanding model is used to recognize the text to be processed to obtain at least one regular text segment and at least one text speech segment in the text to be processed.

[0030] Step S130: If the to-be-processed voice is played, the enhanced voice segment is played in a manner different from that of playing the regular voice segment.

[0031] During implementation, in order to emphasize the playback of the enhanced voice segment, the regular voice segment and the enhanced voice segment may be played in different ways.

[0032] In the embodiment of the present application, the speech to be processed is first acquired; then, the speech to be processed is recognized using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment from the speech to be processed; finally, when the speech to be processed is played, the enhanced speech segment is played in a different manner from the regular speech segment. In this way, by playing the regular speech segment and the enhanced speech segment in different manners, the user can quickly grasp the key content in a large text paragraph or the key speech in a speech segment.

[0033] In some embodiments, before step S120 of "using a semantic understanding model to recognize the speech to be processed to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed", a method for training the speech understanding model is further included, including the following steps:

[0034] Step 121: Construct a speech data set;

[0035] Here, the speech data set includes a certain number of speech samples. Users can determine the speech samples according to recognition requirements or obtain speech samples from a speech sample library. During implementation, speech samples can also be increased or decreased according to changes in requirements.

[0036] Step 122: performing enhanced speech annotation on the speech data set to obtain a fully annotated speech data set;

[0037] During the implementation process, each speech sample in the speech data needs to be enhanced with speech annotation, that is, the speech segments that need to be enhanced in the speech sample need to be annotated. After the enhanced speech annotation is completed, the annotated speech sample can be obtained, that is, the annotated speech dataset can be obtained.

[0038] Step 123: Use the annotated speech data set to train the semantic understanding model to obtain a semantic understanding model for identifying enhanced speech segments.

[0039] During the implementation process, each speech sample can be input into the semantic understanding model, and the semantic understanding model can be trained based on the annotation of the speech samples to obtain a semantic understanding model that can recognize and enhance speech segments.

[0040] In this embodiment, a speech dataset is first constructed; then, the speech dataset is annotated with enhanced speech to obtain a fully annotated speech dataset; and finally, the semantic understanding model is trained using the annotated speech dataset to obtain a semantic understanding model for identifying enhanced speech segments. In this way, the semantic understanding model can be trained, and the resulting semantic understanding model can be used to identify enhanced speech segments in speech.

[0041] Figure 2A A speech processing method provided in an embodiment of the present application is as follows: Figure 2A As shown, the following steps are included:

[0042] Step S210: Acquire the speech to be processed;

[0043] Step S220: using a semantic understanding model to recognize the speech to be processed, and obtaining at least one regular speech segment and at least one enhanced speech segment in the speech to be processed;

[0044] Step S230: determining a play time period of the at least one enhanced speech segment in the speech to be processed;

[0045] During implementation, the play start time and end time of the enhanced voice segment in the speech to be processed may be identified to determine the play time period of the enhanced voice segment in the speech to be processed.

[0046] Figure 2B A schematic diagram of playing a voice to be processed provided in an embodiment of the present application is shown as follows: Figure 2B As shown, the schematic diagram includes: the time 21 for playing the voice to be processed and the time 22 for playing the text content, wherein,

[0047] Playing time of the voice to be processed 21, used to display the playing time of the voice to be processed;

[0048] Play text content 22, used to display the text content corresponding to the voice to be processed.

[0049] Figure 2C A schematic diagram of playing enhanced voice segments provided in an embodiment of the present application is shown as follows: Figure 2C As shown, the schematic diagram includes: the time 21 for playing the voice to be processed and the time 22 for playing the text content, wherein,

[0050] The time 21 for playing the voice to be processed includes four time periods, namely 2S (seconds) 211, 2S (seconds) 212, 3S (seconds) 213 and 4S (seconds) 214.

[0051] Playing text content 22 includes displaying text content corresponding to the voice to be processed, wherein the text content corresponding to the four time periods of playing enhanced voice segments is enhanced.

[0052] Comparison Figure 2B and 2C It can be seen that it is easier for users to quickly grasp the key content (enhanced voice clip) in a large paragraph of text, and they can choose to play any of the four time periods of the enhanced voice clip to play the corresponding key content.

[0053] from Figure 2B As can be seen from the figure, based on determining the playing time period of at least one enhanced voice segment in the voice to be processed, the playing time period corresponding to the at least one enhanced voice segment can be displayed.

[0054] Step S240: If the speech to be processed is played, based on the playing time period of the enhanced speech segment in the speech to be processed, the enhanced speech segment is played in a manner different from that of playing the regular speech segment.

[0055] During implementation, since the start time and end time of playing the enhanced voice segment in the voice to be processed are determined, the voice within this time period can be played in a different manner from that of playing regular voice segments.

[0056] In the embodiment of the present application, the playback time period of the at least one enhanced speech segment within the speech to be processed is first determined. Then, when the speech to be processed is played, the enhanced speech segment is played in a manner different from that used to play the regular speech segments, based on the playback time period of the enhanced speech segment within the speech to be processed. In this way, the enhanced speech segment can be played in a manner different from that used to play the regular speech segments, based on the playback time period of the enhanced speech segment.

[0057] In some embodiments, the above step S120 of "using the semantic understanding model to recognize the speech to be processed to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed" can be implemented by the following steps:

[0058] Step 121: using a speech recognition model to convert the speech to be processed into text information;

[0059] Here, the speech recognition model may be a model for converting speech into text information, that is, by outputting the speech to be processed to the speech recognition model, text information corresponding to the speech to be processed may be obtained.

[0060] Step 122: using the semantic understanding model to identify the text information, and obtaining at least one enhanced text segment in the text information;

[0061] Here, the semantic understanding model can be a model for recognizing text information. That is, the semantic understanding model can be used to recognize input text information and obtain at least one enhanced text segment from the text information. For example, a text information including time, place, and person can be input into the semantic understanding model to obtain the time text, place text, and person text from the text information.

[0062] Step 123: Determine the at least one enhanced speech segment based on the at least one enhanced text segment.

[0063] During implementation, since this information is obtained by converting the speech to be processed, the obtained enhanced text can also correspond to two enhanced speech segments of the speech to be processed.

[0064] In this embodiment of the present application, the speech to be processed is first converted into text information using a speech recognition model; then, the text information is recognized using the semantic understanding model to obtain at least one enhanced text segment in the text information; and finally, based on the at least one enhanced text segment, the at least one enhanced speech segment is determined. In this way, the speech recognition model and the semantic understanding model can be used to obtain at least one enhanced text segment and the enhanced speech segment corresponding to the enhanced text segment.

[0065] Figure 3 A method for determining a key speech is provided in an embodiment of the present application, such as Figure 3 As shown, the following steps are included:

[0066] Step S310: Acquire long speech;

[0067] Step S320: using a speech recognition model to recognize the long speech and obtain a text corresponding to the long speech;

[0068] During implementation, a speech recognition model can be used to recognize long speech into text.

[0069] Step S330: using a semantic understanding model to recognize the text and obtain key speech;

[0070] During the implementation process, the key speech can be identified based on the accent of the speech; the speech understanding model can also be used to recognize the text and obtain the key text.

[0071] Step S340: When playing the long speech, perform voice or text emphasis processing on the key speech.

[0072] If the user needs to play the long speech in voice form, the key speech will be emphasized (the volume of the key speech will be amplified) to the user.

[0073] If the user chooses to recognize it as text, the key text will be emphasized (the key content will be emphasized by making it bold or changing its color) and displayed.

[0074] In this embodiment of the present application, after a user receives a long voice message from the other party, the voice content is recognized and the accented and key parts of the voice are highlighted. At the same time, if the message is recognized as text, the key parts are also emphasized by bolding or changing color. In this way, the user can select the key voice segment to listen to based on the recognized content, or quickly grasp the key content in a long text message after the message is recognized as text.

[0075] In some embodiments, the above step S130 "if playing the speech to be processed, playing the enhanced speech segment in a manner different from playing the regular speech segment" can be implemented by the following steps:

[0076] Step 131: Obtain a play instruction for at least one target enhanced voice segment in the at least one enhanced voice segment;

[0077] In some embodiments, the play instruction can be obtained based on the time period mark of the enhanced voice segment in the voice to be processed, for example, Figure 2C As shown, the play instruction of the target enhanced voice segment in at least one enhanced voice segment can be obtained by the user clicking any time period of the time period for playing the enhanced voice segment. The play instruction of the first enhanced voice segment can be obtained based on the user clicking 2S (seconds) 211.

[0078] In some embodiments, the enhanced voice segments may be identified in other ways to obtain a play instruction based on the identification. For example, at least one enhanced voice segment may be sorted by sequence number, and the sequence number identification may be displayed when the voice duration is displayed, so that a play instruction for the enhanced voice segment may be obtained based on the sequence number identification.

[0079] During implementation, the play instruction of one target enhanced voice segment in the at least one enhanced voice segment may be obtained, or the play instructions of at least two target enhanced voice segments may be obtained simultaneously.

[0080] Step 132: Play the at least one target enhanced voice segment in voice or text based on the play instruction.

[0081] In some embodiments, when a playback instruction for a target enhanced voice segment is received, the corresponding target enhanced voice segment can be played. For example, if a user clicks 211 in the time 21 for playing the pending voice, the corresponding voice segment "Let's go hiking together tomorrow" can be played in voice mode, or the text segment "Let's go hiking together tomorrow" can be displayed in text mode, allowing the user to quickly grasp the key points in a large paragraph of text.

[0082] In some embodiments, when playback instructions for two target enhanced voice segments are obtained simultaneously, the corresponding target enhanced voice segments can be played sequentially. For example, if a user clicks 211 and 212 in time 21 of the voice to be processed, the voice segments corresponding to 211, "Let's go hiking together tomorrow" and 212, "Meet at the east gate of the school," can be played sequentially in voice playback mode, or the text segments "Let's go hiking together tomorrow" and 212, "Meet at the east gate of the school," can be displayed in text mode, allowing the user to quickly grasp the key points of a large paragraph of text.

[0083] In an embodiment of the present application, the user can select a target enhanced voice segment to play based on the recognized text, which solves the problem that after the voice is converted into text, the finger touches the position corresponding to the text, and the voice starts playing from the corresponding position. When operating on the mobile phone, the text is relatively small and the finger can easily touch it by mistake.

[0084] Figure 4 A speech processing method provided in an embodiment of the present application is as follows: Figure 4 As shown, the following steps are included:

[0085] Step S410: Acquire the speech to be processed;

[0086] Step S420: using a semantic understanding model to recognize the speech to be processed, and obtaining at least one regular speech segment and at least one enhanced speech segment in the speech to be processed;

[0087] Step S430: determining at least one accent segment in the speech to be processed;

[0088] In some embodiments, the accent segment can be determined based on the volume of the speech in the to-be-processed speech. For example, the accent segment can be determined based on the decibel level of the speech in the to-be-processed speech. A decibel threshold is determined based on the to-be-processed speech, and a speech segment with a decibel level greater than the decibel threshold is determined as the accent segment.

[0089] In some embodiments, the accent segment can also be determined based on the tone of the speech in the speech to be processed. For example, the basic tone in the speech to be processed can be determined first, and then the sound segment that is different from the basic tone can be determined as the accent segment. If the voice of an adult is mixed with the voice of a child, the voice of the child can be determined as the accent segment. Similarly, if the voice of an adult is mixed with the voice of a child, the voice of the adult can be determined as the accent segment.

[0090] In some embodiments, the accent segment can also be determined based on the timbre of the speech to be processed. For example, the basic timbre of the speech to be processed can be determined first, and then the sound segment that is different from the basic timbre can be determined as the accent segment. For the speech to be processed recorded by two people, the timbre of one person can be determined as the basic timbre, and then the speech of the other person can be determined as the accent segment.

[0091] Step S440: determining the at least one accent segment as the at least one enhanced speech segment;

[0092] Here, the determined enhanced voice segment may be the same voice segment as the enhanced voice segment determined in step S420 , or may be a voice segment different from the enhanced voice segment determined in step S420 .

[0093] During implementation, since step S420 and steps S430 and S440 are all performed to determine the enhanced speech segment, there is no restriction on the order in which steps S420, S430, and S440 are performed. They can be performed simultaneously or sequentially. The resulting enhanced speech segment can be obtained by combining the enhanced speech segment determined in step S420 and the enhanced speech segment determined in steps S430 and S440.

[0094] Step S450: When playing the speech to be processed in text form, setting different text playing modes for the text segments corresponding to the enhanced speech segments, and playing the enhanced speech segments, wherein the text playing modes include at least one of the following: color, brightness, font, size, and thickness;

[0095] During the implementation process, different colors can be set for different text fragments. The brightness, font, size and thickness of the text in the text fragment can also be set based on the different colors. Here, the user can set and combine multiple different text playback methods according to actual needs, and the combination methods will not be repeated here.

[0096] Step S460: When playing the speech to be processed in the form of voice, set a different voice playing mode for each of the enhanced voice segments, and play the enhanced voice segments, wherein the voice playing mode includes at least one of the following: volume, pitch, and timbre.

[0097] During implementation, different volumes can be set for different voice segments, and the volume, pitch, and timbre of the voice in the text segment can be set based on the different volumes. Here, users can set and combine multiple different voice playback methods according to actual needs, and the combination methods will not be described in detail.

[0098] In the embodiment of the present application, at least one accent segment in the speech to be processed is determined, and the at least one accent segment is determined as the at least one enhanced speech segment. In this way, the enhanced speech segment can be effectively determined based on the characteristics of the sound.

[0099] In an embodiment of the present application, when the processed speech is played in text form, different text playback modes are set for the text segments corresponding to the enhanced speech segments, and the enhanced speech segments are played, wherein the text playback mode includes at least one of the following: color, brightness, font, size, and thickness; when the processed speech is played in voice form, different voice playback modes are set for the enhanced speech segments, and the enhanced speech segments are played, wherein the voice playback mode includes at least one of the following: volume, pitch, and timbre. In this way, users can more easily grasp the key points in a large paragraph of text based on different text playback modes or voice playback modes.

[0100] Based on the foregoing embodiments, an embodiment of the present application provides a speech processing device, which includes the modules included, each module includes sub-modules, and can be implemented by a processor in an electronic device; of course, it can also be implemented by a specific logic circuit; in the implementation process, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP) or a field programmable gate array (FPGA), etc.

[0101] Figure 5 A schematic diagram of the structure of the speech processing device provided in the embodiment of the present application is shown in FIG. Figure 5 As shown, the apparatus 500 includes:

[0102] An acquisition module 510 is used to acquire the speech to be processed;

[0103] The recognition module 520 is configured to recognize the speech to be processed by using a semantic understanding model, and obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed;

[0104] The playing module 530 is configured to play the enhanced voice segment in a manner different from that of playing the regular voice segment when playing the voice to be processed.

[0105] In some embodiments, the device further includes a construction module, a labeling module and a training module, wherein the construction module is used to construct a speech data set; the labeling module is used to perform enhanced speech labeling on the speech data set to obtain a completed labeled speech data set; and the training module is used to train the semantic understanding model using the completed labeled speech data set to obtain a semantic understanding model for identifying enhanced speech segments.

[0106] In some embodiments, the apparatus further comprises a first determining module configured to determine a playback time period of the at least one enhanced speech segment within the speech to be processed. The playback module is further configured to, when playing the speech to be processed, play the enhanced speech segment in a manner different from that used to play the regular speech segment based on the playback time period of the enhanced speech segment within the speech to be processed.

[0107] In some embodiments, the recognition module includes a conversion submodule, an identification submodule and a determination submodule, wherein the conversion submodule is used to convert the speech to be processed into text information using a speech recognition model; the identification submodule is used to identify the text information using the semantic understanding model to obtain at least one enhanced text segment in the text information; and the determination submodule is used to determine the at least one enhanced speech segment based on the at least one enhanced text segment.

[0108] In some embodiments, the device further includes a second determination module and a third determination module, wherein the second determination module is used to determine at least one accent segment in the speech to be processed; and the third determination module is used to determine the at least one accent segment as the at least one enhanced speech segment.

[0109] In some embodiments, the playback module includes a first playback submodule and a second playback submodule, wherein the first playback submodule is used to set different text playback modes for the text segments corresponding to each enhanced voice segment when playing the voice to be processed in text form, and play the enhanced voice segment, wherein the text playback mode includes at least one of the following: color, brightness, font, size and thickness; the second playback submodule is used to set different voice playback modes for each enhanced voice segment when playing the voice to be processed in voice form, and play the enhanced voice segment, wherein the voice playback mode includes at least one of the following: volume, pitch and timbre.

[0110] In some embodiments, the playback module includes an acquisition submodule and a third playback submodule, wherein the acquisition submodule is used to obtain the playback instruction of at least one target enhanced voice segment in the at least one enhanced voice segment; the third playback submodule is used to play the at least one target enhanced voice segment in voice playback mode or text mode based on the playback instruction.

[0111] The description of the above device embodiment is similar to the description of the above method embodiment and has similar beneficial effects as the method embodiment. For technical details not disclosed in the device embodiment of this application, please refer to the description of the method embodiment of this application for understanding.

[0112] It should be noted that, in the embodiment of the present application, if the above method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable an electronic device (which can be a mobile phone, tablet computer, laptop computer, desktop computer, etc.) to execute all or part of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program code, such as a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0113] Correspondingly, an embodiment of the present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the speech processing method provided in the above embodiment are implemented.

[0114] Correspondingly, an embodiment of the present application provides an electronic device, Figure 6 A hardware entity diagram of an electronic device provided in an embodiment of the present application, such as Figure 6 As shown, the hardware entity of the device 600 includes: a memory 601 and a processor 602, the memory 601 stores a computer program that can be run on the processor 602, and the processor 602 implements the steps of the speech processing method provided in the above embodiment when executing the program.

[0115] The memory 601 is configured to store instructions and applications executable by the processor 602, and can also cache data to be processed or processed by the processor 602 and various modules in the electronic device 600 (for example, image data, audio data, voice communication data and video communication data), which can be implemented through flash memory (FLASH) or random access memory (RAM).

[0116] It should be noted that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0117] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments.

[0118] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0119] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0120] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0121] In addition, all functional units in the embodiments of the present application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the above-mentioned integrated units can be implemented in the form of hardware or in the form of hardware plus software functional units.

[0122] Those skilled in the art will understand that all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, and other media that can store program codes.

[0123] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application can essentially or in other words be embodied in the form of a software product that contributes to the relevant technology. The computer software product is stored in a storage medium and includes several instructions for enabling an electronic device (which can be a mobile phone, tablet computer, laptop computer, desktop computer, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0124] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0125] The features disclosed in the several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0126] The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0127] The above is merely an embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A speech processing method, comprising: Get the voice to be processed; Recognizing the speech to be processed by using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed; If the to-be-processed speech is played, the enhanced speech segment is played in a manner different from that of playing the regular speech segment; Based on the determination of the playing time period of at least one of the enhanced voice segments in the voice to be processed, displaying the playing time period corresponding to at least one of the enhanced voice segments; In response to clicking to play a target time segment of the at least one play time segment, the enhanced voice segment corresponding to the target time segment is played in a manner different from that of playing the regular voice segment.

2. The method according to claim 1, before using the semantic understanding model to recognize the speech to be processed to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed, the method further comprises: Build a speech dataset; Performing enhanced speech annotation on the speech data set to obtain a fully annotated speech data set; The semantic understanding model is trained using the annotated speech data set to obtain a semantic understanding model for identifying enhanced speech segments.

3. The method according to claim 1, wherein the step of recognizing the speech to be processed using a semantic understanding model to obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed comprises: Using a speech recognition model, converting the speech to be processed into text information; recognizing the text information using the semantic understanding model to obtain at least one enhanced text segment in the text information; The at least one enhanced speech segment is determined based on the at least one enhanced text segment.

4. The method of claim 1, further comprising: determining at least one accent segment in the speech to be processed; The at least one accent segment is determined as the at least one reinforced speech segment.

5. The method according to claim 1, wherein if the to-be-processed speech is played, the enhanced speech segment is played in a manner different from that of the regular speech segment, comprising: When playing the speech to be processed in text form, setting different text playing modes for the text segments corresponding to the enhanced speech segments, and playing the enhanced speech segments, wherein the text playing modes include at least one of the following: color, brightness, font, size, and thickness; When the to-be-processed speech is played in the form of speech, different speech playing modes are set for each of the enhanced speech segments to play the enhanced speech segments, wherein the speech playing mode includes at least one of the following: volume, pitch and timbre.

6. The method according to claim 1, wherein if the to-be-processed speech is played, the enhanced speech segment is played in a manner different from that of the regular speech segment, comprising: Obtaining a play instruction for at least one target enhanced voice segment in the at least one enhanced voice segment; The at least one target enhanced voice segment is played in a voice playback mode or a text playback mode based on the playback instruction.

7. A speech processing device, comprising: An acquisition module is used to acquire the speech to be processed; a recognition module, configured to recognize the speech to be processed by using a semantic understanding model, and obtain at least one regular speech segment and at least one enhanced speech segment in the speech to be processed; A playing module, configured to play the enhanced voice segment in a manner different from that of playing the regular voice segment when playing the voice to be processed; A display module is configured to display a playback time period corresponding to at least one enhanced voice segment based on the determined playback time period of at least one enhanced voice segment in the voice to be processed; The playing module is further configured to play the enhanced voice segment corresponding to the target time period in response to clicking to play a target time period of the at least one playing time period.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, wherein: When the processor executes the program, the steps in the method according to any one of claims 1 to 6 are implemented.

9. A storage medium, characterized in that: Executable instructions are stored, which are used to cause a processor to execute the steps of the method according to any one of claims 1 to 6.

Images