A parameter adjustment method and device

Abstract

The present application discloses a parameter adjustment method and device, the method comprising: obtaining first speech data, and performing speech segmentation on the first parameter to obtain a speech segmentation result; performing speech recognition processing on the speech segmentation result to obtain a processed text; determining adjustment data of the first parameter based on the processed text; and adjusting the first parameter based on the adjustment data. The above implementation scheme uses speech recognition technology to perform speech recognition on the speech obtained by segmentation based on the first parameter to obtain a processed text. By analyzing the content of the processed text, it can be determined whether the segmented speech has the problem of not being segmented in detail or being segmented in detail, so that the first parameter is adjusted in real time according to the analysis result to obtain a first parameter that is adaptive to different speech data to meet the user's usage needs.

Description

Technical Field

[0001] The present application relates to audio processing technology, and more specifically, to a parameter adjustment method and device. Background Art

[0002] The VAD (Voice Activity Detection) module is commonly used in the front end of call systems and speech recognition systems. It can detect in real time whether the audio signal contains speech. Subsequent modules can optimize performance based on the VAD detection results.

[0003] When people speak, there is usually a pause between sentences. Based on this feature, VAD will split a long speech into multiple short speech according to the pause duration. Reasonable segmentation of long speech is an important performance indicator of VAD.

[0004] Generally, VAD will segment long speech according to a fixed pause duration threshold in the application, that is, when the pause duration between two speech segments is greater than the pause duration threshold, the two speech segments are segmented; and when the pause duration between two speech segments is less than the pause duration threshold, the two speech segments are output as a whole speech segment. However, different subjects have different speaking habits and speaking speeds, and a fixed pause duration threshold cannot be adapted to the segmentation of all speech. Summary of the invention

[0005] In view of this, this application provides the following technical solutions:

[0006] A parameter adjustment method, comprising:

[0007] Obtaining first speech data, and performing speech segmentation on the first speech data based on the first parameter to obtain a speech segmentation result;

[0008] Performing speech recognition processing on the speech segmentation result to obtain a processed text;

[0009] determining adjustment data of the first parameter based on the processed text;

[0010] The first parameter is adjusted based on the adjustment data.

[0011] Optionally, performing speech recognition processing on the speech segmentation result to obtain a processed text includes:

[0012] Performing speech recognition on the speech segmentation result to obtain a recognized text;

[0013] The recognized text is subjected to natural language processing to obtain a processed text, wherein the natural language processing includes at least one of homophone correction, sentence segmentation, and punctuation mark addition.

[0014] Optionally, determining the adjustment data of the first parameter based on the processed text includes:

[0015] When the first object in the processed text satisfies a first condition, it is determined to increase the first parameter, wherein the first condition includes that the number of the first objects exceeds a first set value or the proportion of the first objects in the processed text exceeds a second set value.

[0016] Optionally, the first object is a single character or a word or sentence whose number of characters does not exceed a third set value.

[0017] Optionally, determining the adjustment data of the first parameter based on the processed text includes:

[0018] Determining whether there is a target text in the processed text whose length exceeds a fourth set value;

[0019] If so, determine to reduce the first parameter.

[0020] Optionally, after determining that the processed text includes a target text whose length exceeds a fourth set value, the method further includes:

[0021] determining whether the number of designated punctuation marks included in the target text exceeds a fifth set value;

[0022] If it exceeds, the step of determining to adjust the first parameter to a smaller value is entered.

[0023] Optionally, determining the adjustment data of the first parameter based on the processed text includes:

[0024] determining whether the number of designated punctuation marks included in the processed text exceeds a sixth set value;

[0025] If it exceeds, determine to reduce the first parameter.

[0026] Optionally, determining the adjustment data of the first parameter based on the processed text includes:

[0027] determining an adjustment level of the first parameter based on the processed text;

[0028] An adjustment amount for the first parameter is determined based on the adjustment level.

[0029] Optionally, the adjustment data includes an adjustment method, and the adjusting the first parameter based on the adjustment data includes:

[0030] Based on the adjustment method indicated by the adjustment data, the first parameter is adjusted with a set step value.

[0031] A parameter adjustment device, comprising:

[0032] A speech acquisition module, used to obtain first speech data and perform speech segmentation on the first speech data based on the first parameter to obtain a speech segmentation result;

[0033] A speech recognition module, used for performing speech recognition processing on the speech segmentation result to obtain a processed text;

[0034] A parameter determination module, configured to determine adjustment data of the first parameter based on the processed text;

[0035] A parameter adjustment module is used to adjust the first parameter based on the adjustment data.

[0036] Through the above technical solution, it can be known that the embodiment of the present application discloses a parameter adjustment method and device, the method comprising: obtaining first speech data, and performing speech segmentation on the first parameter to obtain a speech segmentation result; performing speech recognition processing on the speech segmentation result to obtain a processed text; determining adjustment data of the first parameter based on the processed text; and adjusting the first parameter based on the adjustment data. The above implementation scheme uses speech recognition technology to perform speech recognition on the speech obtained by segmentation based on the first parameter to obtain a processed text. By analyzing the content of the processed text, it can be determined whether the segmented speech has the problem of not being segmented detailed enough or being segmented too detailed, so that the first parameter is adjusted in real time according to the analysis result to obtain a first parameter that is adaptive to different speech data to meet the user's usage needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without creative work.

[0038] Figure 1 A flow chart of a parameter adjustment method disclosed in an embodiment of the present application;

[0039] Figure 2 A flowchart of a processed text obtained by the process disclosed in the embodiment of the present application;

[0040] Figure 3 A flowchart of determining adjustment data of a first parameter disclosed in an embodiment of the present application;

[0041] Figure 4 A flowchart of another parameter adjustment method disclosed in an embodiment of the present application;

[0042] Figure 5 A flowchart of determining adjustment data of a first parameter disclosed in an embodiment of the present application;

[0043] Figure 6 A schematic diagram of the structure of a parameter adjustment device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0045] The embodiments of the present application can be applied to electronic devices. The present application does not limit the product form of the electronic device, which can include but is not limited to smart phones, tablet computers, wearable devices, personal computers (PCs), netbooks, etc., and can be selected according to application requirements.

[0046] Figure 1 This is a flow chart of a parameter adjustment method disclosed in an embodiment of the present application, see Figure 1 As shown, the parameter adjustment method may include:

[0047] Step 101: Obtain first speech data, and perform speech segmentation on the first speech data based on a first parameter to obtain a speech segmentation result.

[0048] The first voice data may be original voice data that has not been subjected to any segmentation or segmentation processing, or voice data that has only undergone some basic processing such as noise reduction and polishing. The first voice data may be directly collected by a voice collection device, or received from other devices, or read from a storage medium. The embodiment of the present application does not impose a fixed restriction on the method of obtaining the first voice data.

[0049] Speech segmentation is performed based on the first parameter, and the speech can be but is not limited to using VAD (Voice Activity Detection) technology to segment the speech. In the process of implementing speech segmentation, it needs to be implemented based on some parameters, and these parameters may include the first parameter. Specifically, the first parameter can characterize the pause duration of the segmented speech; that is, when the pause duration between two parts of speech exceeds the value of the first parameter, the segmentation action is performed to divide the two parts of speech into two shorter speech segments; and when the pause duration between the two parts of speech is less than or equal to the value of the first parameter, the segmentation action is not performed, and the two parts of speech are output as a whole speech data.

[0050] Step 102: Perform speech recognition processing on the speech segmentation result to obtain a processed text.

[0051] The speech segmentation result may include multiple speech segments, and the durations of different speech segments may be the same or different, that is, the multiple speech segments in the speech segmentation result are speech segments with different durations.

[0052] After obtaining the speech segmentation result, each speech segment contained therein can be subjected to speech recognition by speech recognition technology to obtain the corresponding processed text. The semantic recognition technology can be, but is not limited to, ASR (Automatic Speech Recognition) technology.

[0053] In the following embodiments, the specific implementation of step 102 will be described in detail, and will not be introduced in detail here.

[0054] Step 103: Determine adjustment data of the first parameter based on the processed text.

[0055] By analyzing the processed text accordingly, some data of the content in the processed text that meets the specific conditions can be determined, and these data can reflect whether the content of the processed text is not segmented carefully enough or segmented too trivially. Since the processed text corresponds to the speech segmentation result obtained by segmentation based on the first parameter, the segmentation effect of the processed text can reflect the segmentation effect of the first speech data.

[0056] When it is determined that the segmentation effect of the processed text is not detailed enough, the first parameter needs to be reduced. For example, the initial first parameter is 2 seconds, so that some two parts of speech with a short pause time will not be segmented, and if the first parameter is reduced to 1.5 seconds, some two parts of speech with a pause time in the range of 1.5-2 seconds will be segmented.

[0057] When it is determined that the segmentation effect of the processed text is relatively trivial, the second parameter needs to be increased. For example, the initial first parameter is 1 second, so that two parts of speech with a long pause time between some intervals will not be segmented. If the first parameter is increased to 1.5 seconds, two parts of speech with a pause time between some intervals in the range of 1-1.5 seconds will be segmented.

[0058] It should be noted that the adjustment data of the first parameter may only indicate the adjustment method, such as increasing or decreasing, or the adjustment parameter may also include a specific adjustment value, such as increasing by 0.5 seconds, increasing by 0.3 seconds, decreasing by 1 second, etc.

[0059] Step 104: Adjust the first parameter based on the adjustment data.

[0060] After the adjustment data is determined, the first parameter may be adjusted directly according to the adjustment data, or the first parameter may be adjusted according to a certain rule based on the adjustment data.

[0061] The parameter adjustment method described in this embodiment uses speech recognition technology to perform speech recognition on the speech obtained by segmentation based on the first parameter to obtain a processed text. By analyzing the content of the processed text, it can be determined whether the segmented speech has the problem of not being segmented detailed enough or being segmented too detailed, so that the first parameter is adjusted in real time according to the analysis result to obtain a first parameter that is adaptive to different speech data to meet the user's usage needs.

[0062] Figure 2 The flowchart of the processing text obtained by the processing disclosed in the embodiment of this application is combined with Figure 2 As shown, the speech segmentation result is subjected to speech recognition processing to obtain a processed text, which may include but is not limited to:

[0063] Step 201: Perform speech recognition on the speech segmentation result to obtain a recognized text.

[0064] The recognized text obtained by performing speech recognition on the speech segmentation results only contains text content and does not contain any punctuation marks. In addition, there may be some recognition errors in the recognized text obtained by speech recognition. Therefore, it is necessary to perform further natural language processing on the obtained recognized text to obtain a text result with higher accuracy.

[0065] Step 202: Perform natural language processing on the recognized text to obtain a processed text, wherein the natural language processing includes at least one of homophone correction, sentence segmentation, and punctuation mark addition.

[0066] For example, the accurate content corresponding to the original first voice data is "Factors affecting the commodity market include unit price, quality, sales, etc.". Since the speaker will pause when introducing each factor, the voice "danjia" corresponding to the unit price is segmented out separately, and the recognized text obtained by voice recognition is "stretcher". Through natural language processing, combined with the voice content near the word (factors affecting the commodity market), it can be determined that "stretcher" is an incorrect text and should be "unit price."

[0067] For another example, due to the fast speaking speed of the speaker, the content that should be two sentences may be segmented into one speech segment. For example, the correct text corresponding to the user's speech content is "Are you almost done? I want to go eat." Due to the fast speaking speed, the two sentences "Are you almost done? I want to go eat" are not segmented. However, through natural language processing, it can be determined that "Are you almost done? I want to go eat" are actually two sentences, and the sentence segmentation and corresponding punctuation marks will be added. The processed text content is "Are you almost done? I want to go eat."

[0068] It should be noted that the natural language processing of the recognized text is also performed on a speech segment basis. Although the natural language processing includes sentence segmentation and punctuation, it is only performed within the scope of the text corresponding to the speech segment. The processing will not merge the texts corresponding to two speech segments or split the text of a speech segment.

[0069] This embodiment describes in detail the process of obtaining the processed text to help technical personnel in the field better understand the specific implementation of the present application solution.

[0070] In one implementation, the determining of the adjustment data of the first parameter based on the processed text may include: determining to increase the first parameter when the first object in the processed text satisfies a first condition, wherein the first condition includes that the number of the first object exceeds a first set value or the proportion of the first object in the processed text exceeds a second set value. The first object may be a single word or a word or sentence whose number of words does not exceed a third set value.

[0071] For example, for the sentence "I want to go to the library today", it should be in one speech segment, but if the pause duration represented by the first parameter is short and the speaking speed of the speaker is also slow, the text corresponding to the divided speech segment includes five text contents: "I", "today", "want", "go" and "library", which only contain independent characters and words. Therefore, it can be determined that the speech segmentation is too trivial and the first parameter is too small, and the first parameter needs to be increased.

[0072] In addition to judging the segmentation effect by the number of first objects, in another implementation, the segmentation effect can also be determined by the proportion of the first object in the processed text. Using the above example of "I want to go to the library today", among the 5 text contents, there are 3 separate words, accounting for 60%. Therefore, it can be determined that the speech segmentation is too trivial, the first parameter is too small, and the first parameter needs to be increased.

[0073] In another implementation, determining the adjustment data of the first parameter based on the processed text may include: determining whether there is a target text in the processed text whose length exceeds a fourth set value; if so, determining to reduce the first parameter.

[0074] It is understandable that if the processed text contains a long text, it means that it contains many short voices, and these voices are not segmented because the first parameter is relatively large. Therefore, in this case, the first parameter needs to be reduced.

[0075] For example, when processing a text containing the content "It is snowing heavily outside the window, and this is the first snow of this year! The students in the classroom are no longer focused on the classroom, and they can't wait to go out to play snowball fights and build snowmen. Finally, the bell for the end of get out of class rang, and the children ran out in a swarm. Look, the children on the playground are having so much fun!", and the number of characters in this large paragraph exceeds the set value of 30, it can be determined that a speech segment in the speech segmentation result contains multiple short speech sounds, and the first parameter needs to be reduced.

[0076] Figure 3 This is a flowchart of determining adjustment data of a first parameter disclosed in an embodiment of the present application, see Figure 3 As shown, the determining the adjustment data of the first parameter based on the processed text may include:

[0077] Step 301: Determine whether there is a target text in the processing text whose length exceeds a fourth set value.

[0078] Step 302: If yes, determine whether the number of designated punctuation marks included in the target text exceeds a fifth set value.

[0079] The designated punctuation mark may be a period, an exclamation mark, a question mark, an ellipsis, or other mark that can represent a complete sentence. If the processed text itself is very long, contains a large number of characters, and includes multiple complete sentences, it can be determined that it contains multiple short voices, and the first voice data needs to be segmented more carefully.

[0080] Step 303: If exceeded, determine to reduce the first parameter.

[0081] In this implementation, after determining that the processed text contains a target text whose length exceeds the fourth set value, it is further determined whether the number of specified punctuation marks contained in the target text exceeds the fifth set value, so that the judgment of the processed text is more detailed and reasonable, avoiding the occurrence of overly long sentences in some special contexts and thus the occurrence of misjudgment.

[0082] Figure 4 This is a flowchart of another parameter adjustment method disclosed in the embodiment of the present application, see Figure 4 As shown, the parameter adjustment method may include:

[0083] Step 401: Obtain first speech data, and perform speech segmentation on the first speech data based on a first parameter to obtain a speech segmentation result.

[0084] Step 402: Perform speech recognition processing on the speech segmentation result to obtain a processed text, and proceed to step 403 or step 404.

[0085] Step 403 : When the first object in the processed text satisfies the first condition, determine to increase the first parameter and proceed to step 405 .

[0086] The first condition includes that the number of the first objects exceeds a first set value or the proportion of the first objects in the processed text exceeds a second set value.

[0087] Step 404 : When the number of designated punctuation marks included in the processed text exceeds a sixth set value, determine to reduce the first parameter and proceed to step 405 .

[0088] Step 405: Adjust the first parameter based on the adjustment data.

[0089] This embodiment provides a relatively complete specific implementation scheme, and introduces the conditions under which it is determined whether to increase or decrease the first parameter. Therefore, the first parameter can be adjusted in real time according to the analysis results of the processed text to obtain the first parameter that is adaptive to different voice data to meet the user's usage needs.

[0090] Figure 5 This is a flowchart of determining the adjustment data of the first parameter disclosed in the embodiment of the present application, see Figure 5 As shown, the determining the adjustment data of the first parameter based on the processed text may include:

[0091] Step 501: Determine an adjustment level of the first parameter based on the processed text.

[0092] Processing text can reflect whether the speech segmentation is reasonable, and can also reflect the degree of unreasonable speech segmentation. When the speech segmentation is less trivial, the first parameter can be increased by one level; when the speech segmentation is more trivial, the first parameter can be increased by two levels. The same is true for reducing the first parameter.

[0093] For example, in the sentence "I want to go to the library today", if the processed text is "I today", "want to go" and "library", the first parameter can be increased by the first value; if the processed text is "I", "today", "day", "want", "go", "library", the first parameter can be increased by the second value. The first value is smaller than the second value.

[0094] Step 502: Determine an adjustment amount of the first parameter based on the adjustment level.

[0095] Based on the foregoing, it can be understood that the higher the adjustment level, the greater the adjustment amount of the first parameter.

[0096] This embodiment provides a specific implementation method for determining the adjustment data of the first parameter based on the processed text. This method divides the adjustment work of the first parameter into different levels according to the analysis of the processed text, and determines the adjustment amount of the first parameter according to the adjustment level. This method can quickly adjust the first parameter to a reasonable value, speed up the progress of the rationalization of speech segmentation work, and improve user experience.

[0097] In another implementation, the adjustment data includes an adjustment method, and the adjusting the first parameter based on the adjustment data may include: adjusting the first parameter with a set step value based on the adjustment method indicated by the adjustment data.

[0098] In this implementation, a fixed step value is set for adjusting the first parameter. Each time the first parameter is adjusted, including increasing the first parameter or decreasing the first parameter, the first parameter will be adjusted according to the set step value. This method ensures the standardization of the adjustment of the first parameter, and is particularly suitable for scenarios where the first parameter is adjusted in a small range, and can make the adjustment result more accurate and reasonable.

[0099] In combination with the above content, the core idea of ​​the present application is: the short speech obtained after segmentation by voice activity detection technology is input into the speech recognition system, which will be recognized as text, and then after text post-processing, the text is segmented and punctuation is added to obtain the processed text; and the processed text can reflect whether the speech segmentation is reasonable. For example, if there are more individual words in the processed text result, it means that the speech segmentation is more fragmented, which indicates that the pause duration threshold used for speech segmentation may be set too small; and if there are many words in the processed text result, and it contains multiple symbols representing complete sentences (periods, exclamation marks, question marks, etc.), this indicates that the speech data input into the speech recognition system contains multiple short speech, and the speech segmentation fails to cut these short speech, which indicates that the pause duration threshold of speech segmentation is set too large. Therefore, the first parameter representing the pause duration can be feedback-adjusted according to the result of the text processing, so as to achieve adaptive adjustment of the first parameter.

[0100] For the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0101] The method is described in detail in the embodiments disclosed in the above-mentioned application. The method of the application can be implemented by various forms of devices. Therefore, the application also discloses a device, and a specific embodiment is given below for detailed description.

[0102] Figure 6 This is a schematic diagram of the structure of a parameter adjustment device disclosed in an embodiment of the present application, see Figure 6 As shown, the parameter adjustment device 60 may include:

[0103] The speech acquisition module 601 is used to obtain first speech data and perform speech segmentation on the first speech data based on the first parameter to obtain a speech segmentation result.

[0104] The speech recognition module 602 is used to perform speech recognition processing on the speech segmentation result to obtain a processed text.

[0105] The parameter determination module 603 is used to determine adjustment data of the first parameter based on the processed text.

[0106] The parameter adjustment module 604 is configured to adjust the first parameter based on the adjustment data.

[0107] The parameter adjustment device described in this embodiment uses speech recognition technology to perform speech recognition on the speech obtained by segmentation based on the first parameter to obtain a processed text. By analyzing the content of the processed text, it can be determined whether the segmented speech has the problem of not being segmented detailed enough or being segmented too detailed, so that the first parameter is adjusted in real time according to the analysis result to obtain a first parameter that is adaptive to different speech data to meet the user's usage needs.

[0108] In one implementation, the speech recognition module can be specifically used to: perform speech recognition on the speech segmentation result to obtain a recognized text; perform natural language processing on the recognized text to obtain a processed text, wherein the natural language processing includes at least one of homophone correction, sentence segmentation, and punctuation addition.

[0109] In one implementation, the parameter determination module can be used to determine that the first parameter should be increased when the first object in the processed text satisfies a first condition, wherein the first condition includes that the number of the first objects exceeds a first set value or the proportion of the first objects in the processed text exceeds a second set value.

[0110] In one implementation, the first object is a single character or a word or sentence whose number of characters does not exceed a third set value.

[0111] In one implementation, the parameter determination module may be used to: determine whether there is a target text in the processed text whose length exceeds a fourth set value; if so, determine to reduce the first parameter.

[0112] In one implementation, the parameter determination module may also determine whether the number of specified punctuation marks contained in the target text exceeds a fifth set value after determining that the processed text contains a target text whose length exceeds a fourth set value; and if so, determine to adjust the first parameter to a smaller value.

[0113] In one implementation, the parameter determination module may be used to: determine whether the number of designated punctuation marks included in the processed text exceeds a sixth set value; if so, determine to reduce the first parameter.

[0114] In one implementation, the parameter determination module may be configured to: determine an adjustment level of the first parameter based on the processed text; and determine an adjustment amount of the first parameter based on the adjustment level.

[0115] In one implementation, the adjustment data includes an adjustment method, and the parameter adjustment module can be used to: adjust the first parameter with a set step value based on the adjustment method indicated by the adjustment data.

[0116] Any of the parameter adjustment devices described in the above embodiments includes a processor and a memory. The speech acquisition module, speech recognition module, parameter determination module, parameter adjustment module, etc. in the above embodiments are all stored in the memory as program modules, and the processor executes the above program modules stored in the memory to implement corresponding functions.

[0117] The processor includes a kernel, which retrieves the corresponding program module from the memory. One or more kernels can be set, and the processing of the access data can be realized by adjusting the kernel parameters.

[0118] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0119] An embodiment of the present application provides a storage medium on which a program is stored. When the program is executed by a processor, the parameter adjustment method described in the above embodiment is implemented.

[0120] An embodiment of the present application provides a processor, which is used to run a program, wherein the parameter adjustment method described in the above embodiment is executed when the program is running.

[0121] Furthermore, this embodiment provides an electronic device, including a processor and a memory. The memory is used to store executable instructions of the processor, and the processor is configured to execute the parameter adjustment method described in the above embodiment by executing the executable instructions. The executable instructions include: obtaining first speech data, and performing speech segmentation on the first parameter to obtain a speech segmentation result; performing speech recognition processing on the speech segmentation result to obtain a processed text; determining adjustment data of the first parameter based on the processed text; and adjusting the first parameter based on the adjustment data.

[0122] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0123] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0124] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0125] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A parameter adjustment method, comprising: Obtaining first speech data, and performing speech segmentation on the first speech data based on the first parameter to obtain a speech segmentation result; Performing speech recognition processing on the speech segmentation result to obtain a processed text; Analyze the processed text to determine relevant data of content in the processed text that meets specific conditions, and determine adjustment data of the first parameter based on the relevant data of the content that meets the specific conditions; adjusting the first parameter based on the adjustment data; The relevant data can reflect the degree of segmentation of the first voice data, and the relevant data includes at least one of the following: the number of first objects in the processed text or their proportion in the processed text, wherein the first objects are individual characters or words or sentences whose number of characters does not exceed a third set value; The length of the target text in the processed text, the target text being the text in the processed text whose length exceeds a fourth set value; the number of designated punctuation marks contained in the target text in the processed text; The number of specified punctuation marks contained in the processed text.

2. According to the parameter adjustment method of claim 1, the step of performing speech recognition processing on the speech segmentation result to obtain a processed text comprises: Performing speech recognition on the speech segmentation result to obtain a recognized text; The recognized text is subjected to natural language processing to obtain a processed text, wherein the natural language processing includes at least one of homophone correction, sentence segmentation, and punctuation mark addition.

3. The parameter adjustment method according to claim 1, wherein the step of analyzing the processed text to determine relevant data of content in the processed text that meets specific conditions, and determining the adjustment data of the first parameter based on the relevant data of the content that meets the specific conditions comprises: When the first object in the processed text satisfies a first condition, it is determined to increase the first parameter, wherein the first condition includes that the number of the first objects exceeds a first set value or the proportion of the first objects in the processed text exceeds a second set value.

4. The parameter adjustment method according to claim 1, wherein the step of analyzing the processed text to determine relevant data of content in the processed text that meets specific conditions, and determining the adjustment data of the first parameter based on the relevant data of the content that meets the specific conditions comprises: Determining whether there is a target text in the processed text whose length exceeds a fourth set value; If so, determine to reduce the first parameter.

5. The parameter adjustment method according to claim 4, after determining that the processing text contains a target text whose length exceeds the fourth set value, further comprising: determining whether the number of designated punctuation marks included in the target text exceeds a fifth set value; If it exceeds, the step of determining to adjust the first parameter to a smaller value is entered.

6. The parameter adjustment method according to claim 1, wherein the step of analyzing the processed text to determine relevant data of content in the processed text that meets specific conditions, and determining the adjustment data of the first parameter based on the relevant data of the content that meets the specific conditions comprises: determining whether the number of designated punctuation marks included in the processed text exceeds a sixth set value; If it exceeds, determine to reduce the first parameter.

7. The parameter adjustment method according to claim 1, further comprising: determining an adjustment level of the first parameter based on the processed text; An adjustment amount for the first parameter is determined based on the adjustment level.

8. The parameter adjustment method according to claim 1, wherein the adjustment data includes an adjustment method, and the adjusting the first parameter based on the adjustment data comprises: Based on the adjustment method indicated by the adjustment data, the first parameter is adjusted with a set step value.

9. A parameter adjustment device, comprising: A speech acquisition module, used to obtain first speech data and perform speech segmentation on the first speech data based on the first parameter to obtain a speech segmentation result; A speech recognition module, used for performing speech recognition processing on the speech segmentation result to obtain a processed text; A parameter determination module, configured to analyze the processed text, determine relevant data of content in the processed text that meets specific conditions, and determine adjustment data of the first parameter based on the relevant data of the content that meets the specific conditions; A parameter adjustment module, configured to adjust the first parameter based on the adjustment data; The relevant data can reflect the degree of segmentation of the first voice data, and the relevant data includes at least one of the following: the number of first objects in the processed text or their proportion in the processed text, wherein the first objects are individual characters or words or sentences whose number of characters does not exceed a third set value; The length of the target text in the processed text, the target text being the text in the processed text whose length exceeds a fourth set value; the number of designated punctuation marks contained in the target text in the processed text; The number of specified punctuation marks contained in the processed text.

Images