Speech recognition method, device and equipment and readable storage medium

Abstract

The invention provides a voice recognition method, device and equipment and a readable storage medium, and relates to the technical field of voice recognition, and the method comprises the steps: obtaining historical audio data; obtaining the processed voice data; performing feature extraction to obtain a first voice feature and a second voice feature; when the similarity is greater than a first set threshold value, inputting preset target audio data into a preprocessing model for preprocessing to obtain target voice data; and obtaining a speech recognition result. According to the method, the historical voice data carrying the noise are input into the model for preprocessing, the processed data and the historical voice data not carrying the noise are subjected to feature comparison, and the parameters and the algorithm of the model are continuously optimized, so that the preprocessing model obtained by training can more effectively recognize and remove noise interference, and the noise interference can be effectively eliminated. A clearer and more accurate input signal is provided for a subsequent voice recognition model, it is ensured that only high-quality voice data enter a recognition process, and the recognition accuracy of the voice recognition model is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of speech recognition technology, and in particular to a speech recognition method, device, equipment and readable storage medium. Background Art

[0002] In the existing technology, noise is a key factor affecting the accuracy and robustness of speech recognition systems. Many speech recognition models perform poorly when processing noisy audio data, resulting in increased recognition error rates. Especially in noisy environments, noise interference in speech signals may mask speech information and affect feature extraction and model performance. Therefore, there is an urgent need for an effective method to process noisy speech data to improve the accuracy and robustness of speech recognition systems. Summary of the invention

[0003] The purpose of the present invention is to provide a speech recognition method, device, equipment and readable storage medium to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0004] In a first aspect, the present application provides a speech recognition method, comprising:

[0005] Acquire historical audio data, wherein the historical audio data includes historical speech data with noise and historical speech data without noise;

[0006] Inputting the historical speech data with noise into a preset processing model for preprocessing to obtain processed speech data;

[0007] Performing feature extraction on the processed speech data and the historical speech data without noise to obtain a first speech feature and a second speech feature respectively;

[0008] Calculating the similarity between the first speech feature and the second speech feature, and when the similarity is greater than a first set threshold, inputting the preset target audio data into the preprocessing model for preprocessing to obtain target speech data, where the target audio data is the audio data to be recognized;

[0009] The target speech data is input into a preset speech recognition model for speech recognition to obtain a speech recognition result.

[0010] In a second aspect, the present application also provides a speech recognition device, comprising:

[0011] A first acquisition unit, configured to acquire historical audio data, wherein the historical audio data includes historical speech data with noise and historical speech data without noise;

[0012] A first input unit is used to input historical speech data carrying noise into a preset processing model for preprocessing to obtain processed speech data;

[0013] An extraction unit, used to extract features from the processed speech data and the historical speech data without noise, to obtain a first speech feature and a second speech feature respectively;

[0014] A first calculation unit is used to calculate the similarity between the first speech feature and the second speech feature, and when the similarity is greater than a first set threshold, input the preset target audio data into the preprocessing model for preprocessing to obtain target speech data, where the target audio data is the audio data to be recognized;

[0015] The second input unit is used to input the target voice data into a preset voice recognition model for voice recognition to obtain a voice recognition result.

[0016] In a third aspect, the present application further provides a speech recognition device, comprising:

[0017] Memory for storing computer programs;

[0018] A processor is used to implement the steps of the speech recognition method when executing the computer program.

[0019] In a fourth aspect, the present application further provides a readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned speech recognition method are implemented.

[0020] The beneficial effects of the present invention are:

[0021] The present invention preprocesses historical speech data carrying noise by inputting it into the model, and compares the features of the processed data with the historical speech data without noise, and continuously optimizes the parameters and algorithms of the model, so that the trained preprocessing model can more effectively identify and remove noise interference, provide clearer and more accurate input signals for the subsequent speech recognition model, ensure that only high-quality speech data enters the recognition process, and significantly improve the recognition accuracy of the speech recognition model.

[0022] Other features and advantages of the present invention will be set forth in the following description, and in part will become apparent from the description, or may be understood by practicing embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 Schematic diagram of the speech recognition method described in an embodiment of the present invention;

[0025] Figure 2 is a parameter mapping diagram described in an embodiment of the present invention;

[0026] Figure 3 This is a flow chart of speech recognition model training described in an embodiment of the present invention;

[0027] Figure 4 Schematic diagram of the structure of the speech recognition device described in an embodiment of the present invention;

[0028] Figure 5 Schematic diagram of the structure of the speech recognition device described in the embodiment of the present invention.

[0029] Markings in the figure: 10, first acquisition unit; 20, first input unit; 30, extraction unit; 40, first calculation unit; 50, second input unit; 800, speech recognition device; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0031] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0032] Embodiment 1:

[0033] This embodiment provides a speech recognition method.

[0034] See also Figure 1 , the figure shows that the method includes step S10, step S20, step S30, step S40 and step S50.

[0035] Step S10. Acquire historical audio data, where the historical audio data includes historical speech data with noise and historical speech data without noise;

[0036] Specifically, the historical voice data with noise is the audio data collected by a voice collection device in a normal environment, and the historical audio data will contain noise in the environment and irrelevant human voices; the historical voice data without noise is the data obtained after filtering, denoising and manual listening adjustment of the historical voice data with noise, and it is believed that the historical voice data without noise only contains the required human voice data.

[0037] Step S20. Inputting the historical speech data with noise into a preset processing model for preprocessing to obtain processed speech data;

[0038] Specifically, the historical speech data carrying noise is input into the processing model for preprocessing to obtain processed speech data. By comparing the correlation between the speech features of the processed speech data and the historical speech data without noise, it is determined whether the processing model can perform better denoising on the speech data carrying noise information.

[0039] Step S30. Extract features from the processed speech data and the historical speech data without noise to obtain a first speech feature and a second speech feature respectively;

[0040] Specifically, step S30 specifically includes step S31, step S32, step S33, step S34, step S35, step S36 and step S37:

[0041] Step S31. Calculate the energy value of the processed voice data in each preset time period to obtain multiple energy values;

[0042] Step S32. When the energy value is less than the preset energy threshold, the time period corresponding to the energy value is used as a pause time period;

[0043] Step S33. Divide the processed speech data based on all pause time periods to obtain a plurality of sub-audio data, and calculate the energy value of each sub-audio data to obtain a plurality of target energy values;

[0044] Specifically, due to different speaking habits, different people may speak the same text in different ways, and there may be some incoherence. Therefore, the sentence can be divided into multiple sub-audio data according to the starting and ending points of the speech, that is, the energy value of the corresponding voice segment, so that subsequent processing and analysis are more flexible, and more detailed feature extraction can be performed for different sub-audio data.

[0045] Step S34. Clustering the multiple sub-audio data based on the clustering algorithm and the multiple target energy values ​​to obtain multiple cluster sets, each cluster set containing multiple sub-audio data with small differences in target energy values;

[0046] Specifically, the clustering algorithm may use a distance-based clustering algorithm. This step can more accurately identify data sets with similar features by performing more detailed clustering of the sub-audio data, thereby providing more accurate input for subsequent feature extraction and association analysis.

[0047] Step S35. Calculate the energy threshold range of all sub-audio data contained in each cluster set based on the Laida criterion, and take the cluster set corresponding to the minimum threshold range among all threshold ranges as the abnormal set;

[0048] Step S36. Delete the abnormal sets in all cluster sets to obtain multiple target cluster sets;

[0049] Specifically, by using the Laida criterion and analysis threshold range, abnormal data can be automatically detected and excluded. For example, when normal speech is collected in a relatively quiet environment, the sudden appearance of background noise will cause the energy value of some sub-audio data to increase abnormally. This noise sub-data with a large energy value will interfere with subsequent analysis and affect accuracy. Therefore, after detecting these high-energy noise sub-data, they can be directly deleted to improve the accuracy and reliability of subsequent analysis.

[0050] Step S37: extracting features from the sub-audio data in all target cluster sets, and aggregating the extracted multiple features to obtain a first speech feature;

[0051] Specifically, considering that a target cluster set contains multiple sub-audio data, the speech features of each target cluster set should be composed of all the audio data in the set, which can not only effectively offset the influence of occasional noise in a single sub-audio data, but also ensure that the extracted features are more comprehensive and more representative.

[0052] Specifically, step S37 specifically includes step S371, step S372, step S373, step S374, step S375, step S376 and step S377:

[0053] Step S371. Perform Fourier transform on each sub-audio data, and input the spectrum obtained after Fourier transform into a Mel filter for processing to obtain a plurality of energy logarithms, where the energy logarithm is the logarithm of the energy value of each frequency band output by the Mel filter;

[0054] Step S372: Perform a cosine transform on the energy logarithm to obtain a corresponding plurality of Mel-frequency cepstral coefficients;

[0055] Specifically, by performing Fourier transform and Mel filter processing on each sub-audio data, the spectral characteristics of the audio can be extracted, and the cosine transform further converts the spectral information into Mel-frequency cepstrum coefficients, thereby enhancing the noise resistance and expression ability of the features.

[0056] Step S373. Take the first twelve coefficients of the multiple Mel-frequency cepstral coefficients corresponding to each sub-audio data as characteristic coefficients, and encode all the characteristic coefficients to obtain a characteristic coefficient matrix;

[0057] Specifically, by screening the first twelve main feature coefficients for encoding to form a feature coefficient matrix, the system can focus on the most significant frequency features, lower noise, the impact of speaking style or emotional changes. The feature coefficient matrix obtained after encoding can better represent the speech characteristics of each target cluster set, which helps to improve the accuracy and efficiency of subsequent analysis.

[0058] Step S374. Calculate the correlation values ​​of the two characteristic coefficients in the characteristic coefficient matrix based on the grey correlation analysis algorithm;

[0059] Step S375: passing the information of each feature coefficient to the adjacent feature coefficient, and performing feature update in combination with the information of the adjacent feature coefficient to obtain the updated feature coefficient;

[0060] Specifically, by calculating the correlation between each feature coefficient in the feature coefficient matrix through grey correlation analysis, the intrinsic correlation between features can be revealed, and the information of each feature coefficient is passed to the adjacent features, and the features are updated, so that the features can maintain individual differences while also incorporating the contextual information of adjacent features, thereby improving the overall expressiveness and robustness of the features. This not only optimizes the discriminative power of the features, but also enhances the noise resistance and robustness of the features, providing a more accurate and stable feature description for subsequent speech recognition.

[0061] The calculation formula for feature update is:

[0062]

[0063] Among them, x' s is the updated sth characteristic coefficient; x s is the sth characteristic coefficient; h sz is the correlation value between the sth characteristic coefficient and the adjacent zth characteristic coefficient; x z is the zth characteristic coefficient adjacent to the sth characteristic coefficient; n is the number of characteristic coefficients adjacent to the sth characteristic coefficient.

[0064] Step S376: pooling all updated feature coefficients to obtain multiple local feature information, where the local feature information is obtained by aggregating adjacent updated feature coefficients;

[0065] Step S377. Perform feature splicing on the multiple local feature information to obtain the first speech feature of the processed speech data;

[0066] Specifically, by performing a pooling operation on all updated feature coefficients, the information of adjacent feature coefficients can be aggregated to obtain multiple local feature information, thereby simplifying the feature representation, reducing noise interference while retaining key speech information; then, through feature splicing, multiple local feature information is integrated into the first speech feature, so that the system can form a comprehensive and expressive feature vector, which provides a data basis for subsequent speech analysis and recognition.

[0067] Step S40. Calculate the similarity between the first voice feature and the second voice feature. When the similarity is greater than a first set threshold, input the preset target audio data into the preprocessing model for preprocessing to obtain target voice data, where the target audio data is the audio data to be recognized;

[0068] Specifically, by continuously adjusting the model parameters of the processing model, when the similarity between the features of the processed historical audio data and the features of the historical speech data reaches a first set threshold, it is considered that the current processing model has better noise removal performance and can be used for subsequent preprocessing of speech data to be recognized.

[0069] Specifically, step S40 specifically includes step S41, step S42, step S43, step S44, step S45, step S46, step S47, step S49, step S410 and step S411:

[0070] Step S41. Perform feature conversion on the first speech feature and the second speech feature to obtain a first spectrogram and a second spectrogram respectively;

[0071] Step S42: block processing is performed on the first spectrum graph to obtain a plurality of initial modules;

[0072] Step S43. Calculate the brightness sum of each primary color in the three primary colors included in each initial module to obtain multiple brightness sums, where the brightness sum is the sum of the brightness of each primary color in the initial module;

[0073] Step S44. Determine any brightness sum from a plurality of brightness sums as a first brightness sum;

[0074] Step S45. Determine the second brightness sum located in the neighborhood of the first brightness sum based on a preset neighborhood radius;

[0075] Step S46. When the number of the second brightness sums is not less than a preset minimum number, the first brightness sum and all the second brightness sums are divided into a first target set;

[0076] Step S47. Delete the multiple brightness sums included in the first target set from the multiple brightness sums, and determine any brightness sum therefrom, and repeat the above steps until multiple first target sets are obtained;

[0077] Step S48. Repeat the above steps to determine multiple second target sets corresponding to the second spectrum graph;

[0078] Specifically, the neighborhood radius of the brightness sum is set, that is, the speech energy of the initial module corresponding to all brightness sums in the neighborhood is roughly the same, and a minimum number is set to define the number of brightness sums in the minimum neighborhood of a core point. A first brightness sum is randomly selected, and the difference between the brightness sum and all brightness sums in the target cluster set is calculated. The points within the neighborhood radius are marked as second brightness sums. When the number of second brightness sums is greater than the minimum number, the first brightness sum is the core point, and the first brightness sum and the adjacent second brightness sums are divided into the same target set. The above steps are repeated until all brightness sums are divided into corresponding target sets, thereby determining all target sets, and determining the characteristic information of speech in units of target sets.

[0079] Step S49. Determine a first correlation based on a difference between the number of the first target set and the number of the second target set;

[0080] Specifically, by dividing the neighborhood as described above, the difference between the spectrograms corresponding to the first speech feature and the second speech feature, that is, the difference in speech features, can be effectively compared.

[0081] The calculation formula for the first correlation is:

[0082]

[0083] Wherein, S1 is the first correlation; β is the first control parameter; D1 is the number of the first target set; D2 is the number of the second target set;

[0084] The control parameter is used to control how fast the correlation decreases.

[0085] Step S410. Determine a second correlation based on a difference between the sum of brightness in the first target set and the sum of brightness in the second target set;

[0086] Specifically, step S410 specifically includes step S4101, step S4102, step S4103, step S4104, step S4105, step S4106 and step S4107:

[0087] Step S4101. Determine a partial set from the plurality of first target sets based on a random function as a third target set;

[0088] Specifically, encoding processing may be performed on all first target sets, each first target set has a corresponding code, a random function is used to randomly determine part of the codes from multiple codes, and the set corresponding to the code is used as the third target set.

[0089] Step S4102. Based on the location of the third target set, determine a fourth target set in the second spectrum graph that is closest to the location of the third target set;

[0090] Specifically, the position coordinates of the third target set in the first spectrum graph and the position coordinates of each second target set in the second spectrum graph are determined, and the distance difference between the position coordinates of the third target set and the position coordinates of each second target set in the second spectrum graph is calculated to determine the minimum distance difference. The second target set corresponding to the minimum distance difference is the fourth target set.

[0091] Step S4103. Calculate the position deviation between each third target set and the corresponding fourth target set to obtain a first deviation, where the position deviation includes a lateral position deviation and a longitudinal position deviation;

[0092] Step S4104. Calculate the difference between the sum of brightness in each third target set and the sum of brightness in the corresponding fourth target set as the second deviation;

[0093] Step S4105. Calculate the sum of the first deviation and the second deviation to obtain the target deviation;

[0094] Step S4106. Calculate the product of the target deviation and the control parameter to obtain the target product;

[0095] Step S4107. Calculate the exponential function value with the natural constant as the base and the target product as the exponent to obtain the second correlation;

[0096] Specifically, the calculation formula of the second correlation is:

[0097]

[0098] Where S2 is the second correlation; γ is the second control parameter; D position is the position deviation; D brightness for brightness and deviation;

[0099] The control parameter is used to control how fast the correlation decreases.

[0100] Step S411. Determine the similarity based on the first correlation and the second correlation;

[0101] Specifically, the similarity calculation formula is:

[0102] S=S1+S2

[0103] Among them, S is the similarity; S1 is the first correlation; S2 is the second correlation.

[0104] Step S50: Input the target speech data into a preset speech recognition model for speech recognition to obtain a speech recognition result;

[0105] Specifically, Figure 3 As shown in the figure, the training process of the speech recognition model relies on the sparrow optimization algorithm. First, the sparrow population is initialized and the fitness is evaluated. The individuals in the sparrows are divided into different species. Then, different species of sparrows are optimized generation by generation by simulating natural selection, crossover and mutation operations until the optimal solution is found, thereby constructing the corresponding speech recognition model.

[0106] Specifically, step S50 specifically includes step S51, step S52, step S53, step S54, step S55, step S56, step S57, step S58, step S59, step S510 and step S511:

[0107] Step S51. Obtaining historical voice text data corresponding to the historical voice data;

[0108] Specifically, the historical voice text data is the corresponding text data that is automatically identified and manually adjusted based on the historical voice data.

[0109] Step S52. Initializing the position of the sparrow population based on the chaotic mapping to obtain a plurality of initial position parameters;

[0110] Specifically, Logistic chaotic mapping is used to initialize the sparrow population, which can traverse the state of the sparrow population without repetition within a certain range, and make the sparrow population relatively evenly distributed in the entire search space, which not only increases the diversity of the initial sparrow population, but also avoids falling into the local optimal situation during the search process of the sparrow algorithm.

[0111] Specifically, step S52 specifically includes step S521, step S522, step S523, step S524 and step S525:

[0112] Step S521. Obtain the initial value and branch parameter of each sparrow in the sparrow population, the initial value is within the first set range, and the branch parameter is within the second set range;

[0113] Step S522. Calculate the product of the initial value and the branch parameter to obtain a first value;

[0114] Step S523. Calculate the difference between the preset value and the initial value to obtain a second value;

[0115] Step S524. Calculate the product of the first value and the second value to obtain a chaotic value;

[0116] Step S525. Repeat the calculation of the product of the initial value and the branch parameter, the calculation of the difference between the preset value and the initial value, and the calculation of the product of the first value and the second value until the number of chaotic values ​​reaches the set number of times, and use all chaotic values ​​as initial position parameters;

[0117] Specifically, the calculation formula for the initial position parameters is:

[0118]

[0119] in, is the chaos value of sparrow i at the t+1th time; is the t-th chaotic value of sparrow i, and its value range is [0,1]; α is the branch parameter that determines whether the chaotic map is in a chaotic state, and its value range is [0,4].

[0120] In this embodiment, the chaotic value after mapping each parameter 2,000 times is used as the initial position parameter of the sparrow. Figure 2 As shown, it is the parameter mapping diagram after two thousand iterations of chaos mapping.

[0121] Step S53. construct a model based on the initial position parameters to obtain an initial recognition model;

[0122] Step S54. Based on the initial position parameters of all sparrows, determine the initial fitness of all sparrows;

[0123] Specifically, assuming that the number of sparrows in the sparrow population is n, and each sparrow individual is a solution in the d-dimensional solution space, the fitness of the sparrow population is:

[0124]

[0125] Among them, F(X) is the fitness of the sparrow population; x n,d is the position of sparrow n in the population in the d-dimensional space; f([x n, 1x n,2 …x n,d ]) is the individual fitness of sparrow n.

[0126] Step S55. Divide all sparrows into discoverers, followers and guards based on the initial position parameters;

[0127] Step S56. Based on a preset position update formula, the position parameters of the discoverer, follower and sentinel are updated respectively to obtain updated position parameters;

[0128] Specifically, dividing sparrows into discoverers, followers, and sentinels helps to simulate the natural selection process and enhance the diversity and adaptability of the population. Dynamically adjusting sparrows with different roles through the position update formula can optimize the search strategy and improve the global search capability, thereby converging to the optimal solution more effectively.

[0129] Taking into account that the global search is too dependent on the position of the discoverer, the spiral upward search method of iterative optimization in the whale optimization algorithm is borrowed, and an adaptive weight factor is introduced. This allows the discoverer in the sparrow algorithm to search with a larger step size in the early stage when exploring the next area after the position is updated, and to perform a small step size convergent exploration in the later stage.

[0130] In the embodiment of the present application, the location update formula of the discoverer is:

[0131]

[0132] in, is the j-th position of sparrow i in the population after the T+1th iteration; is the j-th position of sparrow i in the population after the T-th iteration; δ is the adaptive weight factor; iter maxis the maximum number of iterations of the population; θ is a uniformly distributed random number with a value range of (0, 1]; R2 is the warning value with a value range of [0, 1]; ST is the warning threshold with a value range of [0.5, 1]; Q is a random number that follows a normal distribution; Z is a matrix with 1 row and d columns; is the worst position of sparrow i in the population after the T-th iteration; is the optimal position of sparrow i in the population after the T-th iteration; W(T) is an adaptive weight factor, a function that decreases as the number of iterations T increases; A is the amplitude factor of the spiral search; B is the frequency factor of the spiral search; is the upper bound constraint of sparrow i; is the lower bound constraint of sparrow i; iter i is the current iteration number of sparrow i.

[0133] When R2 < ST, it indicates that when the warning value is lower than the safety value, there is no predator in the foraging environment, and at this time, the discoverer can conduct extensive searches; when R2 ≥ ST, it indicates that some sparrows in the population have discovered the predator and issued warnings to other sparrows, and all sparrows need to quickly fly to a safe area to forage.

[0134] In the embodiments of this application, the position update of the sparrow followers is as follows;

[0135]

[0136] where, is the j-th dimension position of sparrow i in the population after the (T + 1)-th iteration; X w is the position of the worst sparrow in the population after the T-th iteration; is the j-th dimension position of sparrow i in the population after the T-th iteration; Q is a random number that follows a normal distribution; is the position of sparrow p in the population after the (T + 1)-th iteration; K is a uniformly distributed random number with a value range of [-1, 1]; d is the spatial dimension; is the j-th dimension position of sparrow p in the population after the T-th iteration; n is the number of individuals in the sparrow population;

[0137] When i > n / 2, it indicates that the i-th follower has not obtained food and is in a hungry state. At this time, it needs to fly to other places to forage to obtain more energy; when i ≤ n / 2, it indicates that the i-th follower can obtain food and does not need to fly to other places to forage to obtain energy.

[0138] To improve the convergence speed and accuracy of the sparrow algorithm, the step size factor and B are corrected. The correction formula for the step size factor is:

[0139]

[0140] Among them, f g is the fitness of the best sparrow in the current population; f w is the fitness of the worst sparrow in the current population; t is the current iteration number; Si is the initialization parameter; rand is a uniformly distributed random factor; and B are step factors, and the value range of B is [0,1].

[0141] The position update formula of the sparrow sentinel is:

[0142]

[0143] in, is the j-th position of sparrow i in the population after the T+1th iteration; is the j-th position of sparrow i in the population after the T-th iteration; is the position of the best sparrow in the population after the Tth iteration; is the position of the worst sparrow in the population after the Tth iteration; f i is the individual fitness of sparrow i; f w is the fitness of the worst sparrow in the current population; f g is the fitness of the best sparrow in the current population; ε is the minimum constant to prevent the denominator from being zero; and B are step factors, and the value range of B is [0,1];

[0144] When f i ≠f g When , it means that the sparrows are at the edge of the population and are very vulnerable to predators; when f i =f g , it means that the sparrows in the middle of the population are also in danger and need to get closer to other sparrows to reduce the risk of being preyed upon.

[0145] Step S57. Determine the current iteration number based on the location update number;

[0146] Step S58. Calculate the mutation rate based on the preset mutation rate, the preset minimum mutation rate, the preset maximum number of iterations and the current number of iterations;

[0147] Step S59. Determine multiple target sparrows from the sparrow population based on the mutation rate, and optimize the updated position parameters of the multiple target sparrows based on the Cauchy distribution algorithm to obtain optimized position parameters;

[0148] Specifically, the mutation operation can expand the search space of the sparrow population, but it does not require every sparrow individual to perform this operation in each iteration. The target sparrow for mutation needs to be determined through the mutation rate, so as to optimize the position parameters. While appropriately maintaining the diversity of the population, excessive random disturbances can be avoided, thereby improving the convergence speed and search efficiency of the algorithm.

[0149] Step S510. When the number of alerters is less than the second set threshold, a target sparrow is determined from the current sparrow population, as well as the optimized position parameters and initial fitness corresponding to the target sparrow, and the target sparrow is the sparrow with the largest fitness in the sparrow population;

[0150] Step S511. Based on the historical voice data, the historical voice text data, the optimized position parameters of the target sparrow and the initial fitness of the target sparrow, the preset initial recognition model is adjusted and optimized to obtain a voice recognition model;

[0151] Specifically, considering that when the number of alerters is large, it is beneficial to the global search of the algorithm; when the number of alerters is small, it is more conducive to the algorithm to perform a fast local search in a small range. Therefore, in the embodiment of the present application, a higher alerter ratio is set in the early stage of the algorithm to enhance the global exploration ability of the population, and the alerter ratio is gradually reduced as the number of iterations increases, thereby improving the convergence speed of the algorithm.

[0152] The formula for updating the proportion of vigilant is:

[0153]

[0154] Among them, P is the proportion of vigilant people; P0 is the initial proportion of vigilant people; T is the current number of iterations; iter max is the maximum number of iterations of the population; P min It is the preset minimum value of the ratio of alerters;

[0155] When P>P min When P≤P min When the conditions are met, the iteration ends, the optimal position and the best fitness value are obtained from the global, the optimal weights and thresholds of the convolutional neural network are determined, and the optimal weights and thresholds are transmitted back to the convolutional neural network for retraining to obtain the speech recognition model.

[0156] Inputting the target voice data into the speech recognition model can achieve rapid recognition and conversion of voice into text. When used in judicial scenarios, it can help judicial mediators record and process case information more efficiently, thereby improving the quality and efficiency of mediation work.

[0157] Embodiment 2:

[0158] like Figure 4 As shown, this embodiment provides a speech recognition device, the device comprising:

[0159] A first acquisition unit 10 is used to acquire historical audio data, where the historical audio data includes historical speech data with noise and historical speech data without noise;

[0160] The first input unit 20 is used to input the historical speech data with noise into a preset processing model for preprocessing to obtain processed speech data;

[0161] An extraction unit 30 is used to extract features from the processed speech data and the historical speech data without noise, and obtain a first speech feature and a second speech feature respectively;

[0162] A first calculation unit 40 is used to calculate the similarity between the first speech feature and the second speech feature. When the similarity is greater than a first set threshold, the preset target audio data is input into the preprocessing model for preprocessing to obtain target speech data, where the target audio data is the audio data to be recognized;

[0163] The second input unit 50 is used to input the target speech data into a preset speech recognition model for speech recognition to obtain a speech recognition result.

[0164] In a specific embodiment disclosed in the present application, the extraction unit 30 includes:

[0165] A second calculation unit, used to calculate the energy value of the processed voice data in each preset time period to obtain multiple energy values;

[0166] The first unit is used to use the time period corresponding to the energy value as a pause time period when the energy value is less than a preset energy threshold;

[0167] A first dividing unit is used to divide the processed speech data based on all pause time periods to obtain a plurality of sub-audio data, and calculate the energy value of each sub-audio data to obtain a plurality of target energy values;

[0168] A clustering unit, used for clustering the multiple sub-audio data based on a clustering algorithm and multiple target energy values ​​to obtain multiple cluster sets, each cluster set containing multiple sub-audio data with small differences in target energy values;

[0169] The second unit is used to calculate the energy threshold range of all sub-audio data contained in each cluster set based on the Laida criterion, and take the cluster set corresponding to the minimum threshold range among all the threshold ranges as the abnormal set;

[0170] A first deleting unit is used to delete abnormal sets in all cluster sets to obtain multiple target cluster sets;

[0171] The aggregation unit is used to extract features from the sub-audio data in all target cluster sets, and aggregate the extracted multiple features to obtain the first speech feature.

[0172] In a specific embodiment disclosed in the present application, the polymeric unit includes:

[0173] A first transform unit is used to perform Fourier transform on each sub-audio data, and input the spectrum obtained after Fourier transform into a Mel filter for processing to obtain a plurality of energy logarithms, where the energy logarithm is the logarithm of the energy value of each frequency band output by the Mel filter;

[0174] A second transformation unit is used to perform a cosine transformation on the energy logarithm to obtain a corresponding plurality of Mel-frequency cepstrum coefficients;

[0175] The third unit is used to take the first twelve coefficients of the multiple Mel-frequency cepstral coefficients corresponding to each sub-audio data as characteristic coefficients, and encode all the characteristic coefficients to obtain a characteristic coefficient matrix;

[0176] A third calculation unit is used to calculate the correlation value of each pair of characteristic coefficients in the characteristic coefficient matrix based on a grey correlation analysis algorithm;

[0177] A transmission unit, used to transmit information of each feature coefficient to an adjacent feature coefficient, and perform feature update in combination with the information of the adjacent feature coefficient to obtain an updated feature coefficient;

[0178] A pooling unit is used to perform a pooling operation on all updated feature coefficients to obtain a plurality of local feature information, where the local feature information is obtained by aggregating adjacent updated feature coefficients;

[0179] The concatenation unit is used to concatenate multiple local feature information to obtain the first speech feature of the processed speech data.

[0180] In a specific implementation disclosed in the present application, the second input unit 50 includes:

[0181] A second acquisition unit, used to acquire historical voice text data corresponding to the historical voice data;

[0182] An initialization unit, used for initializing the position of the sparrow population based on chaotic mapping to obtain a plurality of initial position parameters;

[0183] A construction unit, used for constructing a model based on initial position parameters to obtain an initial recognition model;

[0184] A first determining unit is used to determine the initial fitness of all sparrows based on the initial position parameters of all sparrows;

[0185] A second division unit is used to divide all sparrows into discoverers, followers and guards based on the initial position parameters;

[0186] An updating unit, used to update the position parameters of the discoverer, the follower and the sentinel respectively based on a preset position updating formula to obtain updated position parameters;

[0187] A second determining unit, configured to determine a current number of iterations based on the number of location updates;

[0188] a fourth calculation unit, configured to calculate the mutation rate based on a preset mutation rate, a preset minimum mutation rate, a preset maximum number of iterations, and a current number of iterations;

[0189] A third determining unit is used to determine a plurality of target sparrows from the sparrow population based on the mutation rate, and optimize the updated position parameters of the plurality of target sparrows based on the Cauchy distribution algorithm to obtain optimized position parameters;

[0190] a fourth determination unit, for determining a target sparrow from the current sparrow population, and an optimized position parameter and an initial fitness corresponding to the target sparrow when the number of alerters is less than a second set threshold, wherein the target sparrow is the sparrow with the largest fitness in the sparrow population;

[0191] The optimization unit is used to adjust and optimize the preset initial recognition model based on the historical voice data, the historical voice text data, the optimized position parameters of the target sparrow and the initial fitness of the target sparrow to obtain the voice recognition model.

[0192] In a specific implementation disclosed in the present application, the initialization unit includes:

[0193] A third acquisition unit is used to acquire an initial value and a branch parameter of each sparrow in the sparrow population, wherein the initial value is within a first set range and the branch parameter is within a second set range;

[0194] A product unit, used to calculate the product of the initial value and the branch parameter to obtain a first value;

[0195] A fifth calculation unit, used for calculating the difference between the preset value and the initial value to obtain a second value;

[0196] a sixth calculation unit, used for calculating the product of the first value and the second value to obtain a chaotic value;

[0197] The first repeating unit is used to repeatedly calculate the product of the initial value and the branch parameter, calculate the difference between the preset value and the initial value, and calculate the product of the first value and the second value until the number of chaotic values ​​reaches the set number of times, and all chaotic values ​​are used as initial position parameters.

[0198] In a specific implementation disclosed in the present application, the first calculation unit 40 includes:

[0199] A conversion unit, used for performing feature conversion on the first speech feature and the second speech feature to obtain a first spectrogram and a second spectrogram respectively;

[0200] A processing unit, used for performing block processing on the first spectrum graph to obtain a plurality of initial modules;

[0201] a seventh calculation unit, configured to calculate the brightness sum of each primary color in the three primary colors included in each initial module, and obtain a plurality of brightness sums, wherein the brightness sum is the sum of the brightness of each primary color in the initial module;

[0202] a fifth determining unit, configured to determine any brightness sum from a plurality of brightness sums as a first brightness sum;

[0203] a sixth determining unit, configured to determine a second brightness sum located within a neighborhood of the first brightness sum based on a preset neighborhood radius;

[0204] A third division unit, configured to divide the first brightness sum and all the second brightness sums into a first target set when the sum of the second brightness sums is not less than a preset minimum number;

[0205] A second deleting unit is used to delete the multiple brightness sums included in the first target set from the multiple brightness sums, and determine any brightness sum therefrom, and repeat the above steps until multiple first target sets are obtained;

[0206] A second repeating unit is used to repeat the above steps to determine a plurality of second target sets corresponding to the second spectrum graph;

[0207] A seventh determining unit, configured to determine a first correlation based on a difference between the number of the first target set and the number of the second target set;

[0208] An eighth determining unit, configured to determine a second correlation based on a difference between a brightness sum in the first target set and a brightness sum in the second target set;

[0209] The ninth determining unit is configured to determine a similarity based on the first correlation and the second correlation.

[0210] In a specific implementation manner disclosed in the present application, the eighth determining unit includes:

[0211] a tenth determining unit, configured to determine a partial set from the plurality of first target sets based on a random function as a third target set;

[0212] an eleventh determining unit, configured to determine, based on a location of the third target set, a fourth target set in the second spectrum graph that is closest to a location of the third target set;

[0213] an eighth calculation unit, configured to calculate a position deviation between each third target set and the corresponding fourth target set to obtain a first deviation, wherein the position deviation includes a lateral position deviation and a longitudinal position deviation;

[0214] a ninth calculation unit, configured to calculate a difference between the brightness sum in each third target set and the brightness sum in the corresponding fourth target set as a second deviation;

[0215] A tenth calculation unit, used for calculating the sum of the first deviation and the second deviation to obtain a target deviation;

[0216] an eleventh calculation unit, used for calculating the product of the target deviation and the control parameter to obtain a target product;

[0217] The twelfth calculation unit is used to calculate the exponential function value with the natural constant as the base and the target product as the exponent to obtain the second correlation.

[0218] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0219] Embodiment 3:

[0220] Corresponding to the above method embodiment, this embodiment further provides a speech recognition device. The speech recognition device described below and the speech recognition method described above can refer to each other.

[0221] Figure 5 8 is a block diagram of a speech recognition device 800 according to an exemplary embodiment. Figure 5 As shown, the speech recognition device 800 may include: a processor 801 and a memory 802. The speech recognition device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0222] The processor 801 is used to control the overall operation of the speech recognition device 800 to complete all or part of the steps in the above-mentioned speech recognition method. The memory 802 is used to store various types of data to support the operation of the speech recognition device 800, and these data may include, for example, instructions for any application or method used to operate on the speech recognition device 800, and application-related data, such as contact data, messages sent and received, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read-Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, which is used to receive external audio signals. The received audio signal may be further stored in the memory 802 or sent through the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules may be keyboards, mice, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the speech recognition device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 may include: Wi-Fi module, Bluetooth module, NFC module.

[0223] In an exemplary embodiment, the speech recognition device 800 can be implemented by one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned speech recognition method.

[0224] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by the processor, the steps of the above-mentioned speech recognition method are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions, and the above-mentioned program instructions can be executed by the processor 801 of the speech recognition device 800 to complete the above-mentioned speech recognition method.

[0225] Embodiment 4:

[0226] Corresponding to the above method embodiment, this embodiment further provides a readable storage medium, and the readable storage medium described below and the speech recognition method described above can refer to each other.

[0227] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the speech recognition method of the above method embodiment.

[0228] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0229] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0230] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A speech recognition method, characterized in that: include: Acquire historical audio data, wherein the historical audio data includes historical speech data with noise and historical speech data without noise; Inputting the historical speech data with noise into a preset processing model for preprocessing to obtain processed speech data; Performing feature extraction on the processed speech data and the historical speech data without noise to obtain a first speech feature and a second speech feature respectively; Calculating the similarity between the first speech feature and the second speech feature, and when the similarity is greater than a first set threshold, inputting the preset target audio data into the preprocessing model for preprocessing to obtain target speech data, where the target audio data is the audio data to be recognized; The target speech data is input into a preset speech recognition model for speech recognition to obtain a speech recognition result.

2. The speech recognition method according to claim 1, characterized in that , feature extraction is performed on the processed speech data and the historical speech data without noise, and the first speech feature and the second speech feature are obtained respectively, including: Calculating the energy value of the processed voice data in each preset time period to obtain multiple energy values; When the energy value is less than a preset energy threshold, the time period corresponding to the energy value is used as a pause time period; Dividing the processed speech data based on all pause time periods to obtain a plurality of sub-audio data, and calculating the energy value of each sub-audio data to obtain a plurality of target energy values; Clustering the multiple sub-audio data based on a clustering algorithm and multiple target energy values ​​to obtain multiple cluster sets, each cluster set containing multiple sub-audio data with small differences in target energy values; Calculate the energy threshold range of all sub-audio data contained in each cluster set based on the Laida criterion, and take the cluster set corresponding to the minimum threshold range among all threshold ranges as the abnormal set; Deleting the abnormal set in all cluster sets to obtain multiple target cluster sets; Feature extraction is performed on the sub-audio data in all target cluster sets, and the extracted multiple features are aggregated to obtain the first speech feature.

3. The speech recognition method according to claim 2, characterized in that , extracting features from the sub-audio data in all target cluster sets, and aggregating the extracted multiple features to obtain the first speech feature, including: Performing Fourier transform on each sub-audio data, and inputting the spectrum obtained after Fourier transform into a Mel filter for processing to obtain a plurality of energy logarithms, where the energy logarithm is the logarithm of the energy value of each frequency band output by the Mel filter; Performing a cosine transform on the energy logarithm to obtain a corresponding plurality of Mel-frequency cepstrum coefficients; Taking the first twelve coefficients of the plurality of Mel-frequency cepstrum coefficients corresponding to each sub-audio data as characteristic coefficients, and encoding all the characteristic coefficients to obtain a characteristic coefficient matrix; Calculate the correlation values ​​of each pair of characteristic coefficients in the characteristic coefficient matrix based on the grey correlation analysis algorithm; The information of each feature coefficient is passed to the adjacent feature coefficients, and the feature is updated by combining the information of the adjacent feature coefficients to obtain the updated feature coefficients; Performing a pooling operation on all updated feature coefficients to obtain a plurality of local feature information, wherein the local feature information is obtained by aggregating adjacent updated feature coefficients; The plurality of local feature information are concatenated to obtain a first speech feature of the processed speech data.

4. The speech recognition method according to claim 1, characterized in that , constructing the speech recognition model, including: Acquire historical voice text data corresponding to the historical voice data; Initialize the position of the sparrow population based on chaotic mapping to obtain multiple initial position parameters; Building a model based on the initial position parameters to obtain an initial recognition model; Based on the initial position parameters of all sparrows, the initial fitness of all sparrows is determined; Based on the initial position parameters, all sparrows are divided into finders, followers and guards; Based on a preset position update formula, the position parameters of the discoverer, the follower and the sentinel are updated respectively to obtain updated position parameters; Determining a current number of iterations based on the number of position updates; The mutation rate is calculated based on a preset mutation rate, a preset minimum mutation rate, a preset maximum number of iterations and the current number of iterations; Determine a plurality of target sparrows from the sparrow population based on the mutation rate, and optimize the updated position parameters of the plurality of target sparrows based on the Cauchy distribution algorithm to obtain optimized position parameters; When the number of the alerters is less than a second set threshold, a target sparrow is determined from the current sparrow population, as well as the optimized position parameters and initial fitness corresponding to the target sparrow, wherein the target sparrow is the sparrow with the largest fitness in the sparrow population; Based on the historical voice data, the historical voice text data, the optimized position parameters of the target sparrow and the initial fitness of the target sparrow, the preset initial recognition model is adjusted and optimized to obtain the voice recognition model.

5. The speech recognition method according to claim 4, characterized in that ,Based on the chaotic mapping, the position of the sparrow population is initialized, and multiple initial position parameters are obtained, including: Acquire an initial value and a branch parameter of each sparrow in a sparrow population, wherein the initial value is within a first set range, and the branch parameter is within a second set range; Calculating the product of the initial value and the branch parameter to obtain a first value; Calculate the difference between the preset value and the initial value to obtain a second value; Calculate the product of the first value and the second value to obtain a chaos value; Repeat the calculation of the product of the initial value and the branch parameter, the calculation of the difference between the preset value and the initial value, and the calculation of the product of the first value and the second value until the number of chaotic values ​​reaches the set number of times, and use all chaotic values ​​as initial position parameters.

6. The speech recognition method according to claim 5, characterized in that ,The calculation formula of initial position parameters is: in, is the chaos value of sparrow i at the t+1th time; is the t-th chaotic value of sparrow i, and its value range is [0,1]; α is the branch parameter that determines whether the chaotic map is in a chaotic state, and its value range is [0,4].

7. The speech recognition method according to claim 4, characterized in that ,The discoverer’s position update formula is: in, is the j-th position of sparrow i in the population after the T+1th iteration; is the j-th position of sparrow i in the population after the T-th iteration; δ is the adaptive weight factor; iter max is the maximum number of iterations of the population; θ is a uniform random number, ranging from (0,1]; R2 is the warning value, ranging from [0,1]; ST is the warning threshold, ranging from [0.5,1]; Q is a random number that obeys the normal distribution; Z is a matrix with 1 row and d columns; is the worst position of sparrow i in the population after the Tth iteration; is the optimal position of sparrow i in the population after the Tth iteration; W(T) is the adaptive weight factor, a function that decreases as the number of iterations T increases; A is the amplitude factor of the spiral search; B is the frequency factor of the spiral search; is the upper limit constraint of sparrow i; is the lower limit constraint of sparrow i; iter i is the current iteration number of sparrow i.

8. A speech recognition device, characterized in that: include: A first acquisition unit, configured to acquire historical audio data, wherein the historical audio data includes historical speech data with noise and historical speech data without noise; A first input unit is used to input historical speech data carrying noise into a preset processing model for preprocessing to obtain processed speech data; An extraction unit, used to extract features from the processed speech data and the historical speech data without noise, to obtain a first speech feature and a second speech feature respectively; A first calculation unit is used to calculate the similarity between the first speech feature and the second speech feature, and when the similarity is greater than a first set threshold, input the preset target audio data into the preprocessing model for preprocessing to obtain target speech data, where the target audio data is the audio data to be recognized; The second input unit is used to input the target voice data into a preset voice recognition model for voice recognition to obtain a voice recognition result.

9. The speech recognition device according to claim 8, characterized in that: The extraction unit comprises: A second calculation unit, used to calculate the energy value of the processed voice data in each preset time period to obtain multiple energy values; The first unit is used to use the time period corresponding to the energy value as a pause time period when the energy value is less than a preset energy threshold; A first dividing unit is used to divide the processed speech data based on all pause time periods to obtain a plurality of sub-audio data, and calculate the energy value of each sub-audio data to obtain a plurality of target energy values; A clustering unit, used for clustering the multiple sub-audio data based on a clustering algorithm and multiple target energy values ​​to obtain multiple cluster sets, each cluster set containing multiple sub-audio data with small differences in target energy values; The second unit is used to calculate the energy threshold range of all sub-audio data contained in each cluster set based on the Laida criterion, and take the cluster set corresponding to the minimum threshold range among all the threshold ranges as the abnormal set; A first deleting unit is used to delete the abnormal set in all cluster sets to obtain multiple target cluster sets; The aggregation unit is used to extract features from the sub-audio data in all target cluster sets, and aggregate the extracted multiple features to obtain the first speech feature.

10. The speech recognition device according to claim 9, characterized in that: The polymeric unit comprises: A first transform unit is used to perform Fourier transform on each sub-audio data, and input the spectrum obtained after Fourier transform into a Mel filter for processing to obtain a plurality of energy logarithms, where the energy logarithm is the logarithm of the energy value of each frequency band output by the Mel filter; A second transformation unit is used to perform a cosine transformation on the energy logarithm to obtain a corresponding plurality of Mel-frequency cepstral coefficients; The third unit is used to take the first twelve coefficients of the multiple Mel-frequency cepstral coefficients corresponding to each sub-audio data as characteristic coefficients, and encode all the characteristic coefficients to obtain a characteristic coefficient matrix; A third calculation unit is used to calculate the correlation value of each pair of characteristic coefficients in the characteristic coefficient matrix based on a grey correlation analysis algorithm; A transmission unit, used to transmit information of each feature coefficient to an adjacent feature coefficient, and perform feature update in combination with the information of the adjacent feature coefficient to obtain an updated feature coefficient; A pooling unit, used for performing a pooling operation on all updated feature coefficients to obtain a plurality of local feature information, wherein the local feature information is obtained by aggregating adjacent updated feature coefficients; The concatenation unit is used to concatenate multiple local feature information to obtain the first speech feature of the processed speech data.

Images