A feature-based intelligent voiceprint recognition method and system

Abstract

The present invention belongs to the technical field of voiceprint recognition and discloses a feature-based intelligent voiceprint recognition method and system; the method comprises: collecting a plurality of voiceprint data samples and preprocessing each voiceprint data sample to obtain a comprehensive voice sample; constructing a voice directed graph model based on the comprehensive voice sample, and identifying a core voice sample set from the voice directed graph model; extracting features from the core voice sample set to obtain #imgabs0# comprehensive feature vectors; obtaining a sample selection strategy using a clustering algorithm based on the comprehensive feature vectors and the core voice sample set; adaptively training a voiceprint recognition model using the sample selection strategy based on the comprehensive feature vectors; and applying the trained voiceprint recognition model to voiceprint recognition after the training is completed, thereby giving full play to the advantages of deep learning in the field of voiceprint recognition, not only improving the speed and efficiency of voiceprint recognition, but also reducing the system's false recognition rate and missed recognition rate.

Description

Technical Field

[0001] The present invention relates to the technical field of voiceprint recognition, and more specifically, to a feature-based intelligent voiceprint recognition method and system. Background Art

[0002] Voiceprint recognition has attracted much attention as a contactless means of biological or object identification; however, existing voiceprint recognition methods still have some shortcomings in practical applications; first, the existing methods have a major bottleneck in obtaining high-quality annotated voiceprint data, which requires collecting biological or object sounds in various environments and manually annotating them by professionals, which is a huge workload and costly; due to the complexity and variability of environmental scenes, it is often difficult to obtain enough voiceprint samples to cover various situations, which seriously restricts the generalization ability of voiceprint recognition models; secondly, the existing voiceprint preprocessing method has rigid parameter settings and cannot adapt to the characteristics of different speech signals. For example, in a noisy indoor environment, due to the interference of background noise, fixed denoising parameters may not be able to effectively eliminate noise; in an outdoor environment, due to the influence of natural noise such as wind, fixed denoising parameters may not be able to effectively eliminate noise. The detection parameters may not be able to accurately detect the key voice parts, which will lead to poor preprocessing effects and affect the subsequent voiceprint recognition accuracy; furthermore, the existing voiceprint feature extraction methods are limited to a single dimension and cannot fully characterize the essential characteristics of the speech signal; they cannot fully characterize the dynamic change process of the speech signal, which will inevitably lead to information loss and affect the performance of voiceprint recognition; in addition, in the process of voiceprint recognition model training, the existing methods usually use full data set training or random sampling to select training samples, lacking an effective sample selection strategy. Full data set training can easily lead to overfitting of the model to certain specific users and poor generalization ability; although random sampling can alleviate the overfitting problem to a certain extent, it does not consider the representativeness and diversity between samples, so it is still difficult to ensure the generalization performance of the model, and the training efficiency is low.

[0003] In view of this, the present invention proposes a feature-based intelligent voiceprint recognition method and system to solve the above problems. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a feature-based intelligent voiceprint recognition method, comprising: S1, collecting a plurality of voiceprint data samples, and preprocessing each voiceprint data sample to obtain a comprehensive voice sample;

[0005] S2. constructing a speech directed graph model based on the comprehensive speech samples, and identifying a core speech sample set from the speech directed graph model;

[0006] S3, extract the features of the core speech sample set and obtain A comprehensive feature vector;

[0007] S4. Based on the comprehensive feature vector and the core speech sample set, a clustering algorithm is used to obtain a sample selection strategy;

[0008] S5. Based on the comprehensive feature vector, the voiceprint recognition model is adaptively trained using the sample selection strategy. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition.

[0009] Furthermore, the pre-processing method includes:

[0010] Each voiceprint data sample is denoised to obtain a denoised voiceprint sample, and the silence segments before and after the denoised voiceprint sample are removed to obtain a standard voiceprint sample. The standard voiceprint sample is pre-emphasized to obtain a comprehensive speech sample.

[0011] Furthermore, the denoising process includes:

[0012] Select a silent segment from the voiceprint data sample and divide the frequency band of the silent segment into M sub-bands, each sub-band contains Nm frequency bins;

[0013] For the sub-bands, calculate the sub-band power spectrum of each frame ;

[0014] ;

[0015] in, is the index of the frame, is the index of the frequency bin, is a time domain series, is the time index, is the time index of exponential decay, is an exponential decay factor, and ; is the length of the preset exponential decay window;

[0016] For the sub-bands, calculate the average of the sub-band power spectra of the previous several frames as the initial noise power spectrum estimate ; and the initial noise power spectrum estimate Make updates;

[0017] Ways to update include:

[0018] Set a low quantile threshold and a high quantile threshold ;

[0019] For each frame of the subband , calculate the power spectrum value of each frequency bin of the frame ;Will With the initial noise power spectrum estimate Make comparisons;

[0020] like ;

[0021] Then update the noise power spectrum estimate of the next frame:

[0022] ;

[0023] like ;

[0024] Then update the noise power spectrum estimate of the next frame:

[0025] ;

[0026] otherwise ;

[0027] Smoothing the updated noise power spectrum estimate to obtain a noise power spectrum density estimate for each frame of each subband;

[0028] Define a Wiener filter and calculate the optimal coefficient of the Wiener filter based on the noise power spectral density estimate and the preset adaptive expected signal-to-noise ratio ;

[0029] ;in, is the adaptive expected signal-to-noise ratio, For the Frequency domain coefficients of the frame; For the The noise power spectral density estimate of the frame;

[0030] ;

[0031] in, is the preset minimum expected signal-to-noise ratio, is the preset maximum expected signal-to-noise ratio, For the The spectral entropy of the frame, is the maximum spectral entropy;

[0032] The voiceprint data samples are divided into frames and Fourier transformed to obtain the frequency domain coefficients of each frame; the frequency domain coefficients of each frame are multiplied by the optimal coefficients of the Wiener filter to obtain the filtered frequency domain coefficients;

[0033] The filtered frequency domain coefficients are inverse Fourier transformed to obtain the enhanced time domain signal of each frame; the enhanced time domain signals of all frames are overlapped and added to reconstruct the complete enhanced speech signal, which is the denoised ripple sample.

[0034] Furthermore, the silent segment is selected in the following manner:

[0035] Divide the voiceprint data sample into several time frames; for each time frame, calculate its short-time energy En= β × ∑ a =0 A-1 [ y ( a ) - μ ] 2 +(1 - β ) × Enp ,in, represents the time domain sample sequence of the time frame, is the index of the sampling point of the time domain sample sequence, is the length of the time frame, is the mean amplitude of the time domain sample sequence of the time frame, is the forgetting factor, and ; is the short-term energy of the previous time frame;

[0036] Define the initial adaptive energy threshold , where is the mean of the short-time energy of the voiceprint data sample; for each time frame, update the adaptive energy threshold of the current time frame to ;

[0037] ;

[0038] in, is the smoothing factor, is the adaptive energy threshold of the previous time frame, is the long-term statistical threshold; traverse all time frames, and if the short-term energy of any time frame is less than the corresponding adaptive energy threshold, it is marked as a silent frame; find the longest segment of continuous silent frames, and record its starting frame number and ending frame number; starting from the starting frame number and ending at the ending frame number, extract the voiceprint data sample corresponding to this segment, which is the silent segment.

[0039] Furthermore, the method of performing the pre-emphasis processing includes:

[0040] Perform pre-emphasis filtering on the standard voiceprint sample to obtain a pre-emphasized voice signal, and normalize the amplitude of the pre-emphasized voice signal to complete the pre-emphasis processing and obtain a comprehensive voice sample;

[0041] The formula for performing pre-emphasis filtering is:

[0042] ;

[0043] in, is the pre-emphasized speech signal, is the index of the time point; is the regulating factor, is the low-frequency component of the standard voiceprint sample, is the high-frequency component of the standard voiceprint sample, For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at ; is the second-order coefficient; is the pre-emphasis intensity control parameter, is the wavelet basis function;

[0044] The method of performing amplitude normalization includes:

[0045] The preset decomposition layer number L1 is used to decompose the pre-emphasized speech signal into L1 wavelet packets to obtain 2 L1 wavelet packet components; calculate the energy of each wavelet packet component, that is, the sum of the squares of all amplitudes, calculate the sum of the energies of all wavelet packet components E_total, calculate the normalized weight W(l') of the l'th wavelet packet component, that is, the ratio of the energy of the corresponding wavelet packet component to the sum of the energies E_total; for the l'th wavelet packet component, obtain its maximum amplitude max_wpl' and minimum amplitude min_wpl'; scale the maximum amplitude max_wpl' and the minimum amplitude min_wpl' to the range of [-W(l'), W(l')] to obtain the normalized l'th wavelet packet component, perform wavelet packet reconstruction on all normalized wavelet packet components, and obtain the final normalized speech signal.

[0046] Furthermore, the method of constructing the speech directed graph model includes:

[0047] Each comprehensive speech sample is divided into frames to obtain a series of speech frames; a feature vector is extracted from each speech frame, where the feature vector is a Mel-frequency cepstral coefficient or a linear prediction coefficient;

[0048] Calculate the similarity of the feature vectors between any two speech frames as the similarity between the speech frames, and construct a weighted undirected graph based on the similarity between the speech frames ,in is a set of speech frames, is a set of edges, and the speech frames are used as nodes of a weighted undirected graph. If the similarity between two speech frames is less than a preset similarity threshold, the two nodes are connected in the weighted undirected graph, and the weight of the edge is defined as the similarity between the two speech frames;

[0049] The formula for calculating similarity is:

[0050] ;

[0051] in, There are two speech frames and The distance measure between Speech frame The short-term energy For another speech frame Short-term energy; and are two adjustment parameters; Two speech frames and similarity between is the direction zero-crossing metric function;

[0052] Directional zero-crossing metric function Where, There are two speech frames and The angle between the eigenvectors of Represents two speech frames and The absolute value of the difference between the zero-crossing rates of is the zero-crossing adjustment parameter;

[0053] Weighted Undirected Graph Perform the transpose operation to obtain its transposed graph ,Will The direction of each edge in is reversed to obtain a speech directed graph model;

[0054] The ways to perform a transpose operation include:

[0055] Construct a graph neural network model, use the graph attention network as the skeleton network of the graph neural network model, and define the input of the graph neural network model as a weighted undirected graph The topological structure and nodes of the weighted undirected graph are output as the low-dimensional continuous embedding representation of the node; for each node, its MFCC feature vector is extracted as the initial node feature; The initial node features are input into the graph neural network model, and the training objective of the graph neural network model is defined as node classification or link prediction;

[0056] The parameters of the graph neural network model are trained by the back-propagation algorithm. The graph neural network model learns the low-dimensional continuous embedding representation of each node. Based on the low-dimensional continuous embedding representation of all nodes, the adjacency matrix A_emb of the embedding space is constructed. The elements of the adjacency matrix A_emb are the Euclidean distances between the low-dimensional continuous embedding representations of the nodes. The adjacency matrix A_emb is transposed to obtain the transposed adjacency matrix A_emb_T. The directed graph is reconstructed based on the transposed adjacency matrix A_emb_T, that is, the transposed graph .

[0057] Furthermore, the recognition method of the core speech sample set includes:

[0058] All nodes in the speech directed graph model are randomly arranged in any area on the two-dimensional plane as the initial state of all nodes;

[0059] For any two nodes and , calculate the repulsive force, attractive force and directional force of the edge between the two;

[0060] For each node , the repulsive force, attractive force and directional force acting on it are vector synthesized to obtain the resultant force ; According to the magnitude and direction of the resultant force, calculate the node displacement;

[0061] The displacement calculation formula is:

[0062] ;

[0063] in, is the displacement scaling factor, For the joint efforts direction, is the preset desired direction, 、 、 and is the displacement adjustment parameter; For nodes The velocity vector of

[0064] All nodes are moved according to the calculated displacement to obtain a new node layout; the energy of the new node layout is calculated. If the energy is less than the preset iteration threshold, the iteration is terminated; at this time, several nodes are gathered in the central area of any area, and the speech frames corresponding to the nodes in the central area are the core speech sample set.

[0065] Furthermore, for each speech frame in the core speech sample set, its short-time square root amplitude, short-time energy, short-time average amplitude and short-time zero-crossing rate are extracted as time domain features;

[0066] For each speech frame in the core speech sample set, its spectral entropy, spectral flatness, spectral median, spectral effective bandwidth and spectral peak frequency are extracted as frequency domain features;

[0067] The extracted time domain features and frequency domain features are concatenated at the vector level in a fixed order to construct the comprehensive feature vector of the corresponding speech frame; that is, A comprehensive feature vector;

[0068] The method for obtaining the sample selection strategy includes:

[0069] Each speech frame in the core speech sample set is taken as a sample, and the comprehensive feature vector corresponding to the sample is taken as a row to construct a The characteristic matrix of is the number of samples, is the dimension of the comprehensive feature vector;

[0070] The preset window radius R is used for each sample. , in is the center and the radius is Within the spherical range, calculate the sample The sample density ;

[0071] ;in, are other samples within the spherical range, is the kernel function, is the volume of the spherical range, is the regularization parameter, for The sample weights of is the bandwidth parameter;

[0072] The samples with a sample density greater than the preset minimum sample density threshold Min are marked as core samples; for any two core samples, if there exists a series of samples such that the distance between any two adjacent core samples is less than R, then the two core samples are density-reachable;

[0073] Construct a density-reachable matrix , where the elements M1[J 1 ,J 1' ]=1 Indicates core sample and core samples Is density-reachable; traverse the density-reachable matrix , all density-reachable core samples are grouped into one cluster;

[0074] For each cluster, calculate the ratio of the number of samples in it to the total number of samples, which is recorded as the strategy ratio ; According to the strategy ratio , randomly select the corresponding samples; the process of randomly selecting samples from each cluster is called the sample selection strategy.

[0075] Furthermore, the method of adaptively training the voiceprint recognition model using the sample selection strategy includes:

[0076] Randomly selected samples from each cluster and their corresponding comprehensive feature vectors are used as training data sets. The samples are annotated according to voiceprint types as labels of the training data sets.

[0077] Define the deep neural network model structure as the basic architecture of the voiceprint recognition model; and initialize the parameters of the voiceprint recognition model;

[0078] The comprehensive feature vectors corresponding to the randomly selected samples in any cluster are taken as a batch to train the voiceprint recognition model, and the back propagation algorithm and the optimization algorithm are used to update the parameters of the voiceprint recognition model. The training loss function is defined, and the performance of the voiceprint recognition model on the batch is evaluated according to the training loss function to determine whether the performance of the voiceprint recognition model on the batch meets the expectations. If so, the next round of iteration is entered, and the voiceprint recognition model is trained using the comprehensive feature vectors corresponding to the randomly selected samples in the next cluster. If not, the parameters of the voiceprint recognition model are adjusted according to the gradient of the value of the training loss function during the training process, and the training is continued; until the number of training rounds reaches the preset training maximum value, the training is stopped to obtain a trained voiceprint recognition model.

[0079] A feature-based intelligent voiceprint recognition system, which is used to implement the feature-based intelligent voiceprint recognition method, includes: a voiceprint collection and processing module, which is used to collect a number of voiceprint data samples, pre-process each voiceprint data sample, and obtain a comprehensive voice sample;

[0080] The graph model construction and screening module builds a speech directed graph model based on comprehensive speech samples and identifies a core speech sample set from the speech directed graph model;

[0081] Feature extraction module is used to extract features from the core speech sample set to obtain A comprehensive feature vector;

[0082] The strategy planning module uses clustering algorithm to obtain sample selection strategy based on comprehensive feature vectors and core speech sample sets;

[0083] The model training and delivery module uses a sample selection strategy to adaptively train the voiceprint recognition model based on the comprehensive feature vector. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition. The various modules are connected through wired and / or wireless means.

[0084] The technical effects and advantages of the feature-based intelligent voiceprint recognition method and system of the present invention are as follows:

[0085] The present invention improves the accuracy and reliability of voiceprint recognition, especially performs well in complex noisy environments, greatly enhances the system's adaptability to different types of sound signals, and enables it to flexibly respond to various voiceprint recognition scenarios; secondly, by extracting comprehensive time domain and frequency domain features, it can more accurately capture the essential characteristics of the speech signal, thereby significantly improving the accuracy of voiceprint recognition; in addition, through optimized sample selection and feature extraction strategies, it can more effectively utilize the key information in the data set, improve the learning efficiency of the model, and its adaptive training mechanism enables the model to be dynamically optimized according to different data characteristics, significantly improving the generalization ability of the system; it is suitable for a wider range of voice scenarios; in addition, it also has strong anti-interference ability and environmental adaptability, and can maintain stable recognition performance under various complex conditions; it not only improves the speed and efficiency of voiceprint recognition, but also reduces the system's false recognition rate and missed recognition rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 A schematic diagram of a feature-based intelligent voiceprint recognition method of the present invention;

[0087] Figure 2 Schematic diagram of a feature-based intelligent voiceprint recognition system of the present invention. DETAILED DESCRIPTION

[0088] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0089] Example 1

[0090] See also Figure 1 As shown, this embodiment provides a feature-based intelligent voiceprint recognition method, including:

[0091] S1. Collect several voiceprint data samples and preprocess each voiceprint data sample to obtain a comprehensive voice sample;

[0092] S2. constructing a speech directed graph model based on the comprehensive speech samples, and identifying a core speech sample set from the speech directed graph model;

[0093] S3, extract the features of the core speech sample set and obtain A comprehensive feature vector;

[0094] S4. Based on the comprehensive feature vector and the core speech sample set, a clustering algorithm is used to obtain a sample selection strategy;

[0095] S5. Based on the comprehensive feature vector, the voiceprint recognition model is adaptively trained using the sample selection strategy. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition.

[0096] Voiceprint data sample collection methods include:

[0097] Use on-site recording equipment to record and collect target sounds in actual environments, such as recording human voices in different situations. It should be noted that during the collection process, the authenticity and diversity of the voice samples should be ensured as much as possible to improve the robustness and accuracy of the voiceprint recognition model.

[0098] The methods for preprocessing each voiceprint data sample include:

[0099] Each voiceprint data sample is denoised to obtain a denoised voiceprint sample. Endpoint detection is performed based on the denoised voiceprint sample, that is, the silent segments (no sound areas) before and after the denoised voiceprint sample are removed to obtain a standard voiceprint sample. The standard voiceprint sample is then pre-emphasized to obtain a comprehensive speech sample.

[0100] The methods for denoising include:

[0101] Select a silent segment from the voiceprint data sample, divide the frequency band of the silent segment into M sub-bands, each sub-band contains Nm frequency bins; sub-bands, calculate the sub-band power spectrum of each frame ;

[0102] ;

[0103] in, is the index of the frame, is the index of the frequency bin, is a time domain sequence (the original voiceprint data sample is a continuous analog signal, which needs to be digitized through sampling to obtain a discrete digital sequence, which is the time domain sequence). is the time index, is the time index of exponential decay, Indicates that the time domain sequence is time-shifted. is an exponential decay factor, and ; The length of the preset exponential decay window.

[0104] For the sub-bands, calculate the average of the sub-band power spectra of the previous several frames as the initial noise power spectrum estimate ; and the initial noise power spectrum estimate to update.

[0105] Specifically, set a low quantile threshold and a high quantile threshold ,like , corresponding to the 20% low quantile; , corresponding to the 80% high quantile.

[0106] For each frame of the subband , calculate the power spectrum value of each frequency bin of the frame ;Will With the initial noise power spectrum estimate Make a comparison.

[0107] like , then update the noise power spectrum estimate of the next frame ;

[0108] like , then update the noise power spectrum estimate of the next frame ;otherwise ;

[0109] The updated noise power spectrum estimate is smoothed, such as by using median filtering or exponential averaging, to obtain a noise power spectrum density estimate for each frame of each subband.

[0110] Define a Wiener filter and calculate the optimal coefficient of the Wiener filter based on the noise power spectral density estimate and the preset adaptive expected signal-to-noise ratio ;

[0111] ;in, is the adaptive expected signal-to-noise ratio, For the Frequency domain coefficients of the frame; For the The noise power spectral density estimate for the frame.

[0112] ;

[0113] in, is the preset minimum expected signal-to-noise ratio, is the preset maximum expected signal-to-noise ratio, For the The spectral entropy of the frame, is the maximum spectral entropy.

[0114] The voiceprint data samples are divided into frames and Fourier transformed to obtain the frequency domain coefficients of each frame; the frequency domain coefficients of each frame are multiplied by the optimal coefficients of the Wiener filter to obtain the filtered frequency domain coefficients.

[0115] The filtered frequency domain coefficients are inverse Fourier transformed to obtain the enhanced time domain signal of each frame; the enhanced time domain signals of all frames are overlapped and added to reconstruct the complete enhanced speech signal, which is the denoised ripple sample.

[0116] Furthermore, the silent segment selection method includes:

[0117] The voiceprint data samples are divided into several time frames. Usually, the frame length of the time frame is 20-30ms and the frame shift is about 10ms. For each time frame, its short-time energy is calculated. .

[0118] En= β × ∑ a =0 A-1 [ y ( a ) - μ ] 2 +(1 - β ) × Enp ,in, represents the time domain sample sequence of the time frame, is the index of the sampling point of the time domain sample sequence, is the length of the time frame, is the mean value (mean value of amplitude) of the time domain sample sequence of the time frame, is the forgetting factor, and ; Used to control the weight of the current frame energy and the previous frame energy; is the short-term energy of the previous time frame.

[0119] Define the initial adaptive energy threshold , where is the mean of the short-term energy of the voiceprint data samples.

[0120] For each time frame, update the adaptive energy threshold of the current time frame to ;

[0121] ;

[0122] in, is the smoothing factor, used to adjust the weight, is the adaptive energy threshold of the previous time frame, It is the long-term statistical threshold, which means the energy threshold obtained based on statistics over a longer time window.

[0123] Traverse all time frames. If the short-term energy of any time frame is less than the corresponding adaptive energy threshold, mark it as a silent frame; find the longest segment of continuous silent frames and record its starting frame number and ending frame number; starting from the starting frame number and ending at the ending frame number, extract the voiceprint data sample corresponding to this segment, which is the silent segment.

[0124] Furthermore, the pre-emphasis processing method includes:

[0125] The standard voiceprint sample is pre-emphasized and filtered to obtain a pre-emphasized voice signal, and the pre-emphasized voice signal is amplitude normalized to complete the pre-emphasis processing and obtain a comprehensive voice sample.

[0126] The formula for pre-emphasis filtering is:

[0127] ;

[0128] in, is the pre-emphasized speech signal (sequence), is the index of the time point; is an adjustment factor used to balance the ratio of low-frequency and high-frequency components. is the low-frequency component of the standard voiceprint sample, is the high-frequency component of the standard voiceprint sample, For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at ; is the second-order coefficient, usually -0.73; is the pre-emphasis intensity control parameter, As the wavelet basis function, a wavelet basis function with good high-pass characteristics can be selected.

[0129] The decomposition layer number L1 is preset, and the pre-emphasized speech signal is decomposed by L1 wavelet packet (select appropriate wavelet basis function, such as Daubechies wavelet, etc.), and 2 L1wavelet packet components; calculate the energy of each wavelet packet component, that is, the sum of the squares of all amplitudes, calculate the sum of the energies of all wavelet packet components E_total, calculate the normalized weight W(l') of the l'th wavelet packet component, that is, the ratio of the energy of the corresponding wavelet packet component to the sum of the energies E_total; for the l'th wavelet packet component, obtain its maximum amplitude max_wpl' and minimum amplitude min_wpl'; scale the maximum amplitude max_wpl' and the minimum amplitude min_wpl' to the range of [-W(l'), W(l')] to obtain the normalized l'th wavelet packet component, perform wavelet packet reconstruction on all normalized wavelet packet components, and obtain the final normalized speech signal.

[0130] This makes the pre-emphasis processing process more intelligent and adaptive, and can dynamically adjust parameters according to the characteristics of different voice signals, thereby achieving better pre-emphasis effects.

[0131] Furthermore, the methods for constructing a speech directed graph model include:

[0132] Each comprehensive speech sample is divided into frames to obtain a series of speech frames; a feature vector is extracted from each speech frame, and the feature vector is a Mel-frequency cepstral coefficient (MFCC) or a linear prediction coefficient (LPC).

[0133] Calculate the similarity of the feature vectors between any two speech frames as the similarity between the speech frames, and construct a weighted undirected graph based on the similarity between the speech frames ,in is a set of speech frames, is a set of edges, and the speech frames are used as nodes of a weighted undirected graph. If the similarity between two speech frames is less than a preset similarity threshold, the two nodes are connected in the weighted undirected graph, and the weight of the edge is defined as the similarity between the two speech frames.

[0134] The formula for calculating similarity is:

[0135] ;

[0136] in, There are two speech frames and The distance metric between them (choose Euclidean distance or Manhattan distance); Speech frame The short-term energy For another speech frame Short-term energy; and There are two adjustment parameters, which are used to control the influence of each item on the similarity. Two speech frames and similarity between is the direction zero-crossing metric function;

[0137] Directional zero-crossing metric function Where, There are two speech frames and The angle between the eigenvectors of Represents two speech frames and The absolute value of the difference between the zero-crossing rates of It is the zero-crossing adjustment parameter; it is used to control the weight of the zero-crossing rate term.

[0138] Weighted Undirected Graph Perform the transpose operation to obtain its transposed graph ,Will The direction of each edge in is reversed to obtain a speech directed graph model.

[0139] The ways to perform a transpose operation include:

[0140] Construct a graph neural network model, use the graph attention network (GAT) as the skeleton network of the graph neural network model, and define the input of the graph neural network model as a weighted undirected graph The topological structure and nodes of the weighted undirected graph are output as the low-dimensional continuous embedding representation of the node; for each node (speech frame), its MFCC feature vector is extracted as the initial node feature; and its initial node features are input into the graph neural network model, and the training objective of the graph neural network model is defined as node classification or link prediction; specifically, the training objective of node classification is to predict which category the node belongs to based on the low-dimensional continuous embedding representation of the node, and a category label is assigned to each node in advance; the model maps the embedding representation of each node to the corresponding category label; by minimizing the loss function of node classification (such as cross entropy loss), the model learns the embedding representation that can well distinguish nodes of different categories.

[0141] The training goal is link prediction, that is, the edge (connection) between nodes is the target to be predicted; the task of the model is to predict whether there is an edge (connection) between two nodes based on the embedding representation of them; specifically, if the similarity between two nodes is high, then there is likely to be an edge between them, conversely, if the similarity between two nodes is low, then there may not be an edge between them; during the training process, the model will learn to map the embedding representation of two nodes to a similarity score; by minimizing the loss function of link prediction (such as binary cross entropy loss), the model can learn an embedding representation that can well predict the connection relationship between nodes.

[0142] The parameters of the graph neural network model are trained by the back-propagation algorithm. The graph neural network model learns the low-dimensional continuous embedding representation of each node. Based on the low-dimensional continuous embedding representation of all nodes, the adjacency matrix A_emb of the embedding space is constructed. The elements of the adjacency matrix A_emb are the Euclidean distances between the low-dimensional continuous embedding representations of the nodes. The adjacency matrix A_emb is transposed to obtain the transposed adjacency matrix A_emb_T. The directed graph is reconstructed based on the transposed adjacency matrix A_emb_T, that is, the transposed graph .

[0143] Furthermore, the recognition methods of the core speech sample set include:

[0144] All nodes (speech frames) in the speech directed graph model are randomly arranged in any area on the two-dimensional plane as the initial state of all nodes; for any two nodes and , calculate the repulsive force, attractive force and directional force of the edge between the two;

[0145] The formula for calculating repulsive force is:

[0146] ;

[0147] in, For two nodes and The repulsive force between is the repulsion coefficient, which is used to adjust the magnitude of the repulsion. is the repulsive force constant, which controls the attenuation rate of the repulsive force, For two nodes and The Euclidean distance between is the repulsion adjustment parameter, which is used to control the influence of similarity on repulsion; For two nodes and The similarity between is calculated in the same way).

[0148] The formula for calculating gravity is:

[0149] ;

[0150] in, For two nodes and The attraction between is the gravitational constant, which controls the decay rate of gravity, For nodes and The weight of the edge between them reflects the degree of connection between the two nodes; For nodes The zero-crossing rate of the corresponding speech frame, For nodes The zero-crossing rate of the corresponding speech frame, Gravity adjustment parameter; used to control the influence of similarity on gravity.

[0151] The calculation formula of the directional force of the edge is:

[0152] ;

[0153] in, For nodes and Directional force between edges; For nodes and The angle between (calculated by substituting the angle of the eigenvector into the calculation); is the hyperbolic tangent function, whose range is (-1, 1); is the similarity influencing parameter, is the distance influence parameter, which is used to control the influence of similarity and distance on the directional force. It is the directional force coefficient; it is used to adjust the magnitude of the directional force.

[0154] For each node , the repulsive force, attractive force and directional force acting on it are vector synthesized to obtain the resultant force ; According to the magnitude and direction of the resultant force, calculate the node displacement (length and direction of movement).

[0155] The displacement calculation formula is:

[0156] ;

[0157] in, is the displacement scaling factor, For the joint efforts direction, is the preset desired direction (such as horizontal direction), 、 、 and It is the displacement adjustment parameter, which controls the influence of several kinds of information on the displacement; For nodes The velocity vector can be expressed as The weighted sum of the displacement vectors in the first few iterations is calculated.

[0158] All nodes are moved according to the calculated displacement to obtain a new node layout; the energy of the new node layout (the sum of the gravitational forces and the gravitational forces between nodes) is calculated. If the energy is less than the preset iteration threshold, the iteration is terminated; at this time, several nodes are gathered in the central area of any area (pre-divided) and are closely connected to each other. The speech frames corresponding to these nodes are the core speech sample set.

[0159] First, perform trial operations on some typical speech samples and draw a curve showing how the node layout energy changes with the number of iterations. When the energy curve tends to be flat and the rate of change is lower than a certain threshold, the threshold can be used as the iteration threshold.

[0160] Furthermore, the feature extraction methods include:

[0161] For each speech frame in the core speech sample set, its short-time square root amplitude, short-time energy, short-time average amplitude and short-time zero-crossing rate are extracted as time domain features.

[0162] For each speech frame in the core speech sample set, its spectral entropy, spectral flatness, spectral median, spectral effective bandwidth and spectral peak frequency are extracted as frequency domain features; the extracted time domain features and frequency domain features are vector-level spliced in a fixed order to construct the comprehensive feature vector of the corresponding speech frame; that is, A comprehensive feature vector.

[0163] Furthermore, the sample selection strategy is obtained by:

[0164] Each speech frame in the core speech sample set is taken as a sample, and the comprehensive feature vector corresponding to the sample is taken as a row to construct a The characteristic matrix of is the number of samples, is the dimension of the comprehensive feature vector.

[0165] Preset window radius , for each sample , in is the center and the radius is Within the spherical range, calculate the sample The sample density ;

[0166] ;in, are other samples within the spherical range, is a kernel function (such as a Gaussian kernel), is the volume of the spherical range, is a regularization parameter used to avoid the situation where the denominator is 0. for The sample weight can be a confidence score, a manually labeled confidence, or a confidence probability predicted by the model. Giving higher-quality samples a greater weight can improve the accuracy of density estimation. is the bandwidth parameter, which controls the sliding window size of the kernel function.

[0167] The samples with a sample density greater than the preset minimum sample density threshold Min are marked as core samples; for any two core samples, if there is a density chain connecting them, that is, there is a series of samples such that the distance between any two adjacent samples is less than R, then the two core samples are density reachable, that is, there is a density chain.

[0168] Construct a density-reachable matrix , where the elements M1[J 1 ,J 1' ]=1 Indicates core sample and core samples Is density-reachable; traverse the density-reachable matrix , all density-reachable core samples are grouped into one cluster.

[0169] For each cluster, calculate the ratio of the number of samples in it to the total number of samples, which is recorded as the strategy ratio ; According to the strategy ratio , randomly select the corresponding samples; the process of randomly selecting samples from each cluster is called the sample selection strategy.

[0170] Specifically, the methods of adaptively training the voiceprint recognition model using the sample selection strategy include:

[0171] The samples randomly selected from each cluster and their corresponding comprehensive feature vectors are used as the training data set, and the samples are annotated according to the voiceprint type (such as wind sound, breaking glass sound, explosion sound, etc.) as the labels of the training data set.

[0172] Define a deep neural network model structure, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a combination of the two, as the basic architecture of the voiceprint recognition model; and initialize the parameters of the voiceprint recognition model, including the number of network layers, the number of nodes, and the activation function.

[0173] The comprehensive feature vectors corresponding to the randomly selected samples in any cluster are taken as a batch to train the voiceprint recognition model, and the backpropagation algorithm and optimization algorithm (such as SGD, Adam, etc.) are used to update the parameters of the voiceprint recognition model, and the training loss function (such as cross entropy) is defined. The performance of the voiceprint recognition model on the batch is evaluated according to the training loss function. The performance is reflected by the accuracy, that is, (the number of correctly predicted samples) / (the total number of samples), to judge whether the performance of the voiceprint recognition model on the batch meets the expectations, that is, the accuracy is greater than the preset accuracy threshold; if it meets the requirements, then enter the next round of iteration and continue to train the voiceprint recognition model using the comprehensive feature vectors corresponding to the randomly selected samples in the next cluster; if it does not meet the requirements, then adjust the parameters of the voiceprint recognition model according to the gradient of the value of the training loss function during the training process, and then continue training; until the number of training rounds reaches the preset training maximum value, stop training, and obtain the trained voiceprint recognition model; apply the trained voiceprint recognition model to the actual voiceprint recognition task; input the voice sample to be recognized, and the voiceprint recognition model will output the corresponding voiceprint type prediction result.

[0174] This embodiment improves the accuracy and reliability of voiceprint recognition, especially in complex noisy environments, greatly enhances the system's adaptability to different types of sound signals, and enables it to flexibly respond to various voiceprint recognition scenarios; secondly, by extracting comprehensive time domain and frequency domain features, it can more accurately capture the essential characteristics of the voice signal, thereby significantly improving the accuracy of voiceprint recognition; in addition, through optimized sample selection and feature extraction strategies, it can more effectively utilize the key information in the data set and improve the learning efficiency of the model. Its adaptive training mechanism enables the model to be dynamically optimized according to different data characteristics, significantly improving the generalization ability of the system; it is suitable for a wider range of voice scenarios; in addition, it also has strong anti-interference ability and environmental adaptability, and can maintain stable recognition performance under various complex conditions; it not only improves the speed and efficiency of voiceprint recognition, but also reduces the system's false recognition rate and missed recognition rate.

[0175] Example 2

[0176] See also Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description of Example 1. A feature-based intelligent voiceprint recognition system is provided, including:

[0177] The voiceprint collection and processing module is used to collect several voiceprint data samples, pre-process each voiceprint data sample, and obtain a comprehensive voice sample;

[0178] The graph model construction and screening module builds a speech directed graph model based on comprehensive speech samples and identifies a core speech sample set from the speech directed graph model;

[0179] Feature extraction module is used to extract features from the core speech sample set to obtain A comprehensive feature vector;

[0180] The strategy planning module uses clustering algorithm to obtain sample selection strategy based on comprehensive feature vectors and core speech sample sets;

[0181] The model training and delivery module uses a sample selection strategy to adaptively train the voiceprint recognition model based on the comprehensive feature vector. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition. The modules are connected through wired and / or wireless means to realize data transmission between modules.

[0182] Example 3

[0183] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the feature-based intelligent voiceprint recognition method provided above is implemented.

[0184] Since the electronic device described in this embodiment is an electronic device used to implement a feature-based intelligent voiceprint recognition method in the embodiment of this application, based on the feature-based intelligent voiceprint recognition method described in the embodiment of this application, those skilled in the art will be able to understand the specific implementation of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of this application will not be described in detail here. As long as those skilled in the art implement the electronic device used in the feature-based intelligent voiceprint recognition method in the embodiment of this application, it falls within the scope of protection of this application.

[0185] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0186] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for users of ordinary skill in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A feature-based intelligent voiceprint recognition method, characterized in that: include: S1. Collect several voiceprint data samples and preprocess each voiceprint data sample to obtain a comprehensive voice sample; S2. constructing a speech directed graph model based on the comprehensive speech samples, and identifying a core speech sample set from the speech directed graph model; The method of constructing the speech directed graph model includes: Each comprehensive speech sample is divided into frames to obtain a series of speech frames; a feature vector is extracted from each speech frame, where the feature vector is a Mel-frequency cepstral coefficient or a linear prediction coefficient; Calculate the similarity of the feature vectors between any two speech frames as the similarity between the speech frames, and construct a weighted undirected graph based on the similarity between the speech frames ,in is a set of speech frames, is a set of edges, and the speech frames are used as nodes of a weighted undirected graph. If the similarity between two speech frames is less than a preset similarity threshold, the two nodes are connected in the weighted undirected graph, and the weight of the edge is defined as the similarity between the two speech frames; The formula for calculating similarity is: ;in, There are two speech frames and The distance measure between Speech frame The short-term energy For another speech frame Short-term energy; and are two adjustment parameters; Two speech frames and similarity between is the direction zero-crossing metric function; Directional zero-crossing metric function Where, There are two speech frames and The angle between the eigenvectors of Represents two speech frames and The absolute value of the difference between the zero-crossing rates of is the zero-crossing adjustment parameter; Weighted Undirected Graph Perform the transpose operation to obtain its transposed graph ,Will The direction of each edge in is reversed to obtain a speech directed graph model; The ways to perform a transpose operation include: Construct a graph neural network model, use the graph attention network as the skeleton network of the graph neural network model, and define the input of the graph neural network model as a weighted undirected graph The topological structure and nodes of the weighted undirected graph are output as the low-dimensional continuous embedding representation of the node; for each node, its MFCC feature vector is extracted as the initial node feature; The initial node features are input into the graph neural network model, and the training objective of the graph neural network model is defined as node classification or link prediction; The parameters of the graph neural network model are trained by the back-propagation algorithm. The graph neural network model learns the low-dimensional continuous embedding representation of each node. Based on the low-dimensional continuous embedding representation of all nodes, the adjacency matrix A_emb of the embedding space is constructed. The elements of the adjacency matrix A_emb are the Euclidean distances between the low-dimensional continuous embedding representations of the nodes. The adjacency matrix A_emb is transposed to obtain the transposed adjacency matrix A_emb_T. The directed graph is reconstructed based on the transposed adjacency matrix A_emb_T, that is, the transposed graph ; S3, extract the features of the core speech sample set and obtain A comprehensive feature vector; S4. Based on the comprehensive feature vector and the core speech sample set, a clustering algorithm is used to obtain a sample selection strategy; S5. Based on the comprehensive feature vector, the voiceprint recognition model is adaptively trained using the sample selection strategy. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition.

2. The feature-based intelligent voiceprint recognition method according to claim 1, characterized in that: The method of performing pretreatment includes: Each voiceprint data sample is denoised to obtain a denoised voiceprint sample, and the silence segments before and after the denoised voiceprint sample are removed to obtain a standard voiceprint sample. The standard voiceprint sample is pre-emphasized to obtain a comprehensive speech sample.

3. The feature-based intelligent voiceprint recognition method according to claim 2, characterized in that: The denoising process includes: Select a silent segment from the voiceprint data sample and divide the frequency band of the silent segment into M sub-bands, each sub-band contains Nm frequency bins; For the sub-bands, calculate the sub-band power spectrum of each frame ; ;in, is the index of the frame, is the index of the frequency bin, is a time domain series, is the time index, is the time index of exponential decay, is an exponential decay factor, and ; is the length of the preset exponential decay window; For the sub-bands, calculate the average of the sub-band power spectra of the previous several frames as the initial noise power spectrum estimate ; and the initial noise power spectrum estimate Make updates; Ways to update include: Set a low quantile threshold and a high quantile threshold ; For each frame of the subband , calculate the power spectrum value of each frequency bin of the frame ;Will With the initial noise power spectrum estimate Make comparisons; like , then update the noise power spectrum estimate of the next frame ; like , then update the noise power spectrum estimate of the next frame ;otherwise ; Smoothing the updated noise power spectrum estimate to obtain a noise power spectrum density estimate for each frame of each subband; Define a Wiener filter and calculate the optimal coefficient of the Wiener filter based on the noise power spectral density estimate and the preset adaptive expected signal-to-noise ratio ; ;in, is the adaptive expected signal-to-noise ratio, For the Frequency domain coefficients of the frame; For the The noise power spectral density estimate of the frame; ;in, is the preset minimum expected signal-to-noise ratio, is the preset maximum expected signal-to-noise ratio, For the The spectral entropy of the frame, is the maximum spectral entropy; The voiceprint data samples are divided into frames and Fourier transformed to obtain the frequency domain coefficients of each frame; the frequency domain coefficients of each frame are multiplied by the optimal coefficients of the Wiener filter to obtain the filtered frequency domain coefficients; The filtered frequency domain coefficients are inverse Fourier transformed to obtain the enhanced time domain signal of each frame; the enhanced time domain signals of all frames are overlapped and added to reconstruct the complete enhanced speech signal, which is the denoised ripple sample.

4. The feature-based intelligent voiceprint recognition method according to claim 3, characterized in that: The method for selecting the silent segment includes: Divide the voiceprint data sample into several time frames; for each time frame, calculate its short-time energy ,in, represents the time domain sample sequence of the time frame, is the index of the sampling point of the time domain sample sequence, is the length of the time frame, is the mean amplitude of the time domain sample sequence of the time frame, is the forgetting factor, and ; is the short-term energy of the previous time frame; Define the initial adaptive energy threshold , where is the mean of the short-time energy of the voiceprint data sample; for each time frame, update the adaptive energy threshold of the current time frame to ; ;in, is the smoothing factor, is the adaptive energy threshold of the previous time frame, is the long-term statistical threshold; traverse all time frames, and if the short-term energy of any time frame is less than the corresponding adaptive energy threshold, it is marked as a silent frame; find the longest segment of continuous silent frames, and record its starting frame number and ending frame number; starting from the starting frame number and ending at the ending frame number, extract the voiceprint data sample corresponding to this segment, which is the silent segment.

5. The feature-based intelligent voiceprint recognition method according to claim 4, characterized in that: The method of performing the pre-emphasis processing includes: Perform pre-emphasis filtering on the standard voiceprint sample to obtain a pre-emphasized voice signal, and normalize the amplitude of the pre-emphasized voice signal to complete the pre-emphasis processing and obtain a comprehensive voice sample; The formula for performing pre-emphasis filtering is: ;in, is the pre-emphasized speech signal, is the index of the time point; is the regulating factor, is the low-frequency component of the standard voiceprint sample, is the high-frequency component of the standard voiceprint sample, For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at For time point The sampling value of the standard voiceprint sample at ; is the second-order coefficient; is the pre-emphasis intensity control parameter, is the wavelet basis function; The method of performing amplitude normalization includes: The preset decomposition layer number L1 is used to decompose the pre-emphasized speech signal into L1 layer wavelet packets to obtain 2 L1 wavelet packet components; calculate the energy of each wavelet packet component, that is, the sum of the squares of all amplitudes, calculate the sum of the energies of all wavelet packet components E_total, calculate the normalized weight W(l') of the l'th wavelet packet component, that is, the ratio of the energy of the corresponding wavelet packet component to the sum of the energies E_total; for the l'th wavelet packet component, obtain its maximum amplitude max_wpl' and minimum amplitude min_wpl'; scale the maximum amplitude max_wpl' and the minimum amplitude min_wpl' to the range of [-W(l'), W(l')] to obtain the normalized l'th wavelet packet component, perform wavelet packet reconstruction on all normalized wavelet packet components, and obtain the final normalized speech signal.

6. The feature-based intelligent voiceprint recognition method according to claim 5, characterized in that: The recognition method of the core speech sample set includes: All nodes in the speech directed graph model are randomly arranged in any area on the two-dimensional plane as the initial state of all nodes; For any two nodes and , calculate the repulsive force, attractive force and directional force of the edge between the two; For each node , the repulsive force, attractive force and directional force acting on it are vector synthesized to obtain the resultant force ; According to the magnitude and direction of the resultant force, calculate the node displacement; The displacement calculation formula is: ;in, is the displacement scaling factor, For the joint efforts direction, is the preset desired direction, 、 、 and is the displacement adjustment parameter; For nodes The velocity vector of All nodes are moved according to the calculated displacement to obtain a new node layout; the energy of the new node layout is calculated. If the energy is less than the preset iteration threshold, the iteration is terminated; at this time, several nodes are gathered in the central area of any area, and the speech frames corresponding to the nodes in the central area are the core speech sample set.

7. The feature-based intelligent voiceprint recognition method according to claim 6, characterized in that: For each speech frame in the core speech sample set, its short-time square root amplitude, short-time energy, short-time average amplitude and short-time zero-crossing rate are extracted as time domain features; For each speech frame in the core speech sample set, its spectral entropy, spectral flatness, spectral median, spectral effective bandwidth and spectral peak frequency are extracted as frequency domain features; The extracted time domain features and frequency domain features are concatenated at the vector level in a fixed order to construct the comprehensive feature vector of the corresponding speech frame; that is, A comprehensive feature vector; The method for obtaining the sample selection strategy includes: Each speech frame in the core speech sample set is taken as a sample, and the comprehensive feature vector corresponding to the sample is taken as a row to construct a The characteristic matrix of is the number of samples, is the dimension of the comprehensive feature vector; The preset window radius R is used for each sample. , in is the center and the radius is Within the spherical range, calculate the sample The sample density ; ;in, are other samples within the spherical range, is the kernel function, is the volume of the spherical range, is the regularization parameter, for The sample weights of is the bandwidth parameter; The samples with a sample density greater than the preset minimum sample density threshold Min are marked as core samples; for any two core samples, if there exists a series of samples such that the distance between any two adjacent core samples is less than R, then the two core samples are density-reachable; Construct a density-reachable matrix , where the elements Indicates core sample and core samples Is density-reachable; traverse the density-reachable matrix , all density-reachable core samples are grouped into one cluster; For each cluster, calculate the ratio of the number of samples in it to the total number of samples, which is recorded as the strategy ratio ; According to the strategy ratio , randomly select the corresponding samples; the process of randomly selecting samples from each cluster is called the sample selection strategy.

8. The feature-based intelligent voiceprint recognition method according to claim 7, characterized in that: The method of adaptively training the voiceprint recognition model using the sample selection strategy includes: Randomly selected samples from each cluster and their corresponding comprehensive feature vectors are used as training data sets. The samples are annotated according to voiceprint types as labels of the training data sets. Define the deep neural network model structure as the basic architecture of the voiceprint recognition model; and initialize the parameters of the voiceprint recognition model; The comprehensive feature vectors corresponding to the randomly selected samples in any cluster are taken as a batch to train the voiceprint recognition model, and the back propagation algorithm and the optimization algorithm are used to update the parameters of the voiceprint recognition model. The training loss function is defined, and the performance of the voiceprint recognition model on the batch is evaluated according to the training loss function to determine whether the performance of the voiceprint recognition model on the batch meets the expectations. If so, the next round of iteration is entered, and the voiceprint recognition model is trained using the comprehensive feature vectors corresponding to the randomly selected samples in the next cluster. If not, the parameters of the voiceprint recognition model are adjusted according to the gradient of the value of the training loss function during the training process, and the training is continued; until the number of training rounds reaches the preset training maximum value, the training is stopped to obtain a trained voiceprint recognition model.

9. A feature-based intelligent voiceprint recognition system, which is used to implement the feature-based intelligent voiceprint recognition method according to any one of claims 1 to 8, characterized in that: include: The voiceprint collection and processing module is used to collect several voiceprint data samples, pre-process each voiceprint data sample, and obtain a comprehensive voice sample; The graph model construction and screening module builds a speech directed graph model based on comprehensive speech samples and identifies a core speech sample set from the speech directed graph model; Feature extraction module is used to extract features from the core speech sample set to obtain A comprehensive feature vector; The strategy planning module uses clustering algorithm to obtain sample selection strategy based on comprehensive feature vectors and core speech sample sets; The model training and delivery module uses a sample selection strategy to adaptively train the voiceprint recognition model based on the comprehensive feature vector. After the training is completed, the trained voiceprint recognition model is applied to voiceprint recognition. The various modules are connected through wired and / or wireless means.

Images