A 2D-TASNet-based machine room in-situ voice noise reduction method

Abstract

The present application belongs to the technical field of machine room noise reduction, and particularly relates to a machine room indoor human voice noise reduction method based on 2D-TASNet, preprocessing: short-time Fourier transform is performed on a mixed signal to convert a one-dimensional time domain signal into a two-dimensional time-frequency spectrum; feature decomposition step: sparse non-negative matrix factorization is used to extract features from the amplitude spectrum of the time-frequency spectrum; feature clustering: the matrix obtained by SNMF decomposition is clustered by K-Means, and the centroid of the pure human voice time-frequency spectrum is used as the clustering center to enhance the robustness of the separation effect; human voice separation step: the original TASNet is adjusted to 2D-TASNet, the features obtained by K-Means clustering are used as input, 2D-TASNet is trained to obtain a target mask, and the separated human voice is reconstructed using the mask. The frequency characteristics of noise and human voice are different, the structure of 2D-TASNet focuses on mining the potential features of various time frequencies, and such a mode can not only utilize the time-frequency characteristics and nonlinear mixing of human voice and noise, but also maintain the precision advantage of the neural network.

Description

Technical Field

[0001] This invention belongs to the field of noise reduction technology for computer rooms, and particularly relates to a method for human voice noise reduction in computer rooms based on 2D-TASNet. Background Technology

[0002] In a data center environment, the continuous heat dissipation of servers generates significant background noise, posing communication challenges for technicians performing daily inspections, installations, debugging, and maintenance. This persistent noise not only hinders effective teamwork but can also damage the hearing of staff. More importantly, the nature of data center work demands highly precise communication, as even minor misunderstandings can lead to equipment malfunctions or other serious consequences. Therefore, developing a technology capable of clearly separating human voices from noise in this unique environment is of paramount importance.

[0003] Traditional signal processing techniques have limitations when dealing with complex, nonlinearly mixed audio signals. For example, they may not be able to effectively remove background noise without sacrificing vocal quality, especially when the noise frequencies are similar to or vary significantly from the vocal frequencies.

[0004] Human voices and background noise (such as server heat dissipation noise) are usually non-linearly mixed, and traditional noise reduction algorithms have difficulty separating them effectively.

[0005] Therefore, this invention starts with the time-frequency spectrum, utilizing the differences in time-frequency characteristics between human voice and noise to achieve their effective separation. Specifically, human voice typically exhibits certain regularity and sparsity in the time-frequency spectrum, while background noise is more random and dispersed. Based on this characteristic, this invention can design algorithms to accurately identify the components of human voice and separate them from complex background noise.

[0006] The two-dimensional TASNet (2D-TASNet) proposed in this invention is used for indoor voice noise reduction in servers. It can accurately identify and separate complex sound signals and provide clear voice output, thereby improving the on-site communication effect for technicians. Summary of the Invention

[0007] The purpose of this invention is to provide a method for noise reduction of human voices in a computer room based on 2D-TASNet, which is used to achieve noise reduction of human voices in the server room based on two-dimensional TASNet (2D-TASNet), thereby accurately identifying and separating complex sound signals, providing clear human voice output, and improving the on-site communication effect of technicians.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0009] A method for human voice noise reduction in a computer room based on 2D-TASNet includes the following steps:

[0010] S1: Preprocessing: Perform a short-time Fourier transform on the mixed time-domain signal of human voice and background noise to convert the one-dimensional time-domain signal into a two-dimensional time-spectrum signal.

[0011] S2: Feature decomposition: Based on the sparsity of human voice, sparse nonnegative matrix decomposition is used to extract features of the input mixed signal;

[0012] S3: Feature clustering: The matrix obtained by decomposing the amplitude spectrum through SNMF needs to be clustered using K-Means;

[0013] S4: Voice Separation: This is an end-to-end separation system that is modified based on the original TASNet. The input of the original TASNet is modified to a two-dimensional amplitude spectrum, and the output is modified to an ideal proportional mask. The parameters and loss function of the internal structure of the network are adjusted to adapt the input and output. The features obtained by K-Means clustering are used as the input of 2D-TASNet to train 2D-TASNet to obtain the target mask.

[0014] S5: Reconstruct the separated human voice using the target mask.

[0015] Preferably, the specific process of step S1 is as follows:

[0016] The time-frequency domain signal is obtained by performing a short-time Fourier transform (STFT) through a Hamming window with a window length of 256 points and a window shift of 128 points. s ( f , t );

[0017] ;

[0018] in, It is a window function. f Represents frequency, s ( t () is a mixed signal of human voice and background noise.

[0019] Preferably, the specific process of feature decomposition in step S2 is as follows:

[0020] S21: Take the modulus value of the two-dimensional time-frequency signal to obtain the amplitude spectrum:

[0021] V( f , t )=| s ( f , t )|;

[0022] S22: The input matrix V is decomposed into an r-dimensional basis matrix W and a coefficient matrix H, with constraints including non-negativity of elements and L1 sparsity constraints, which are reflected in the coefficient matrix H. An SNMF objective function with Euclidean distance is then defined.

[0023] ;

[0024] ;

[0025] in, λ It is a regularization parameter used to control sparsity and reconstruction error; the objective function aims to... H yuan h The value should be as small as possible to ensure the sparsity of the results, where the dimension of V is zero. f * t , W The dimension is f * r , H The dimension is r * t , r For decomposition dimensions.

[0026] Preferably, an iterative process is performed:

[0027] S23: Initialize W and H as random positive matrices, apply sparsity constraints to H, then project each row of H to be non-negative, and set the L1 criterion to achieve the desired sparsity;

[0028] S24: Project each column of W as nonnegative:

[0029] ;

[0030] S25: Project each row of H to be non-negative, and set the unit L1 standard to achieve the desired sparsity:

[0031] ;

[0032] When the cost function converges, exit the loop and output matrices W and H; otherwise, continue iterating.

[0033] Preferably, the specific process of feature clustering in step S3 is as follows:

[0034] S31: The decomposed basis matrix W is a vector set consisting of K basis vectors. The coefficient matrix H consists of K coefficient vectors, and the vector set W consists of K basis vectors, as shown below:

[0035] ;

[0036] ;

[0037] S32: The basis matrix W and coefficient matrix H contain human voice and noise features distributed across different frequency bands, by... w k and h k The submatrix obtained by multiplication F k :

[0038] ;

[0039] submatrix F k These are characteristics of the source signal; therefore, it is necessary to... w k and h k Clustering is performed to obtain the target features. The clustering results in pure features that contain only human voices and no noise.

[0040] Preferably, the specific process of human voice separation in step S4 is as follows:

[0041] S41: The encoder maps the 2D input features of the amplitude spectrum of the mixed signal to a higher-dimensional space.

[0042] ;

[0043] S42: Separation is achieved in high-dimensional space using an estimator;

[0044] S43: The decoder recovers the high-dimensional vector calculated by the estimator into a mask of the same dimension as the encoder input.

[0045] .

[0046] Preferably, the specific process of reconstructing the separated human voice using the target mask in step S5 is as follows:

[0047] After performing a time-spectrum dot product of the mask and the mixed signal, we obtain the time-spectrum containing only the target human voice. Then, we perform an inverse STFT to obtain the separated human voice. The specific formula is as follows:

[0048] ;

[0049] Where s1(t) is the separated human voice signal, M is the estimated mask, S is the time spectrum of the mixed signal, and ISTFT is the inverse Fourier transform.

[0050] The beneficial effects of this invention include:

[0051] This invention provides a 2D-TASNet-based method for human voice noise reduction in computer rooms. The method includes: preprocessing: performing a short-time Fourier transform (STFT) on the mixed signal to convert the one-dimensional time-domain signal into a two-dimensional time-frequency spectrum; feature decomposition: extracting features from the amplitude spectrum of the time-frequency spectrum using sparse nonnegative matrix factorization (SNMF); feature clustering: performing K-Means clustering on the matrix obtained from SNMF decomposition, using the centroid of the pure human voice time-frequency spectrum as the cluster center to enhance the robustness of the separation effect; and human voice separation: adjusting the original TASNet to a 2D-TASNet, using the features obtained from K-Means clustering as input to train the 2D-TASNet to obtain a target mask, and using this mask to reconstruct the separated human voice. Since noise and human voice have different frequency characteristics, the structure of 2D-TASNet focuses on mining various potential time-frequency features. This approach can utilize the respective time-frequency characteristics and nonlinear mixing of human voice and noise while maintaining the accuracy advantage of neural networks.

[0052] First, the time-domain signal is converted into a two-dimensional time-frequency spectrum using short-time Fourier transform, preserving the joint time-frequency characteristics of the signal and providing a richer information foundation for subsequent processing. Then, sparse nonnegative matrix factorization is used to perform feature decomposition on the amplitude spectrum. Leveraging the sparsity of human voice in the time-frequency domain, nonnegativity constraints and L1 sparsity constraints are applied to ensure that the resulting basis matrix and coefficient matrix better reflect the characteristic differences between human voice and noise, effectively separating representative local features and laying a high-quality feature foundation for subsequent clustering and separation.

[0053] Secondly, K-Means clustering is used to cluster the matrix after SNMF decomposition to further filter out pure features containing only human voices and remove noise-related features. This feature purification process reduces the interference of noise on subsequent network training, allowing 2D-TASNet to focus more on learning the feature patterns of human voices and improve the model's accuracy in recognizing and separating human voices.

[0054] Furthermore, the native TASNet is specifically improved by adjusting the input to a two-dimensional amplitude spectrum and the output to an ideally scaled mask, while adapting the network's internal parameters and loss function to better suit the processing requirements of two-dimensional time-frequency features. This improved 2D-TASNet can fully learn the local and global correlations in the time-frequency domain, generating a more accurate target mask, thereby effectively separating human voices from mixed signals. Simultaneously, the end-to-end network design simplifies the process, reduces manual intervention, and improves separation efficiency.

[0055] Furthermore, the background noise in the computer room environment is complex (such as equipment operating noise, electromagnetic interference, etc.). This method can specifically suppress the continuous or sudden noise unique to the computer room by using the sparsity constraint of SNMF and K-Means clustering. Combined with the deep mining capability of 2D-TASNet for two-dimensional time-frequency features, it can still maintain high human voice separation quality even in computer room environments with low signal-to-noise ratio, thus improving the robustness and practicality of the method in real-world scenarios.

[0056] Finally, by performing a dot product operation between the target mask and the time spectrum of the mixed signal, and then reconstructing the human voice using the inverse short-time Fourier transform, noise can be suppressed while preserving the original spectral characteristics and temporal continuity of the human voice to the greatest extent possible. This ensures that the separated human voice is clear and intelligible, and enables effective extraction of human voice signals in computer room scenarios. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the 2D-TASNet-based architecture of the present invention. Detailed Implementation

[0058] The following is in conjunction with the appendix Figure 1 The present invention will be further described in detail below:

[0059] Example 1

[0060] A method for human voice noise reduction in a computer room based on 2D-TASNet includes the following steps:

[0061] S1: Preprocessing: Perform a short-time Fourier transform on the mixed time-domain signal of human voice and background noise to convert the one-dimensional time-domain signal into a two-dimensional time-spectrum signal.

[0062] S2: Feature decomposition: Based on the sparsity of human voice, sparse nonnegative matrix decomposition is used to extract features of the input mixed signal;

[0063] S3: Feature clustering: The matrix obtained by decomposing the amplitude spectrum through SNMF needs to be clustered using K-Means;

[0064] S4: Voice Separation: This is an end-to-end separation system that is modified based on the original TASNet. The input of the original TASNet is modified to a two-dimensional amplitude spectrum, and the output is modified to an ideal proportional mask. The parameters and loss function of the internal structure of the network are adjusted to adapt the input and output. The features obtained by K-Means clustering are used as the input of 2D-TASNet to train 2D-TASNet to obtain the target mask.

[0065] S5: Reconstruct the separated human voice using the target mask.

[0066] In this embodiment, the specific process of step S1 is as follows:

[0067] The time-frequency domain signal is obtained by performing a short-time Fourier transform (STFT) through a Hamming window with a window length of 256 points and a window shift of 128 points. s ( f , t );

[0068] ;

[0069] in, It is a window function. f Represents frequency, s ( t () is a mixed signal of human voice and background noise.

[0070] The specific process of feature decomposition in step S2 is as follows:

[0071] S21: Take the modulus value of the two-dimensional time-frequency signal to obtain the amplitude spectrum:

[0072] V( f , t )=| s ( f , t )|;

[0073] S22: The input matrix V is decomposed into an r-dimensional basis matrix W and a coefficient matrix H, with constraints including non-negativity of elements and L1 sparsity constraints, which are reflected in the coefficient matrix H. An SNMF objective function with Euclidean distance is then defined.

[0074] ;

[0075] ;

[0076] in, λ It is a regularization parameter used to control sparsity and reconstruction error; the objective function aims to... H yuan h The value should be as small as possible to ensure the sparsity of the results, where the dimension of V is zero. f * t , W The dimension is f * r , H The dimension is r * t , r For decomposition dimensions.

[0077] Execute the iterative process:

[0078] S23: Initialize W and H as random positive matrices, apply sparsity constraints to H, then project each row of H to be non-negative, and set the L1 criterion to achieve the desired sparsity;

[0079] S24: Project each column of W as nonnegative:

[0080] ;

[0081] S25: Project each row of H to be non-negative, and set the unit L1 standard to achieve the desired sparsity:

[0082] ;

[0083] When the cost function converges, exit the loop and output matrices W and H; otherwise, continue iterating.

[0084] Example 2

[0085] Based on Example 1, the specific process of feature clustering in step S3 is as follows:

[0086] S31: The decomposed basis matrix W is a vector set consisting of K basis vectors. The coefficient matrix H consists of K coefficient vectors, and the vector set W consists of K basis vectors, as shown below:

[0087] ;

[0088] ;

[0089] S32: The basis matrix W and coefficient matrix H contain human voice and noise features distributed across different frequency bands, by... w k and h k The submatrix obtained by multiplication F k :

[0090] ;

[0091] submatrix F k These are characteristics of the source signal; therefore, it is necessary to... w k and h k Clustering is performed to obtain the target features. The clustering results in pure features that contain only human voices and no noise.

[0092] The clustering process can be described as follows: Because the initial centers of K-Means are random, which can easily lead to unstable results, this invention uses several pure human voice amplitude spectra and calculates their average to obtain a template feature. The centroid of this template feature is used as one of the cluster centers of K-Means to improve the stability of clustering. Next, a KS clustering algorithm based on cosine similarity is applied to each vector until all vectors are clustered. By superimposing these clusters, the sub-features belonging to the human voice class are obtained, thus yielding the target features.

[0093] See Figure 1 2D-TASNet consists of three parts: Encoder: an encoder that maps 2D input features (amplitude spectrum of the mixed signal) to a higher-dimensional space. In this invention, the encoder layer is essentially a convolutional layer, whose function is to map the amplitude spectrum to a high-dimensional learnable embedding space.

[0094] Estimator: An estimator that performs separation in high-dimensional space. The estimator is a fully convolutional structure, composed of multiple convolutional groups, activation functions, and normalization functions stacked together. The core of the estimator is a Temporal Convolutional Network (TCN) module, consisting of eight exponentially increasing expansion factors and Depth-Conv blocks. As the dimension *dim* increases, the expansion factor within each depth convolutional block doubles. Each depth convolutional block has two outputs, called the residual path and the jump connection path. The residual path is used as the input to the next block, and the outputs of all jump connection paths are summed to ensure a sufficiently large temporal context window to capture the long-range correlations of the input signal.

[0095] Decoder: The purpose of a decoder is to recover a high-dimensional vector calculated by the estimator into a mask of the same dimension as the encoder input. The decoder is essentially the inverse of the encoder, that is, a deconvolutional layer with the same parameters as the encoder.

[0096] The specific process of voice separation in step S4 is as follows:

[0097] S41: The encoder maps the 2D input features of the amplitude spectrum of the mixed signal to a higher-dimensional space.

[0098] ;

[0099] S42: Separation is achieved in high-dimensional space using an estimator;

[0100] S43: The decoder recovers the high-dimensional vector calculated by the estimator into a mask of the same dimension as the encoder input.

[0101] .

[0102] Preferably, the specific process of reconstructing the separated human voice using the target mask in step S5 is as follows:

[0103] After performing a time-spectrum dot product of the mask and the mixed signal, we obtain the time-spectrum containing only the target human voice. Then, we perform an inverse STFT to obtain the separated human voice. The specific formula is as follows:

[0104] ;

[0105] Where s1(t) is the separated human voice signal, M is the estimated mask, S is the time spectrum of the mixed signal, and ISTFT is the inverse Fourier transform.

[0106] The mixed signal is preprocessed by performing a short-time Fourier transform to convert the one-dimensional time-domain signal into a two-dimensional time-frequency spectrum. The feature decomposition step involves extracting features from the amplitude spectrum of the time-frequency spectrum using sparse non-negative matrix factorization. The feature clustering step involves performing K-Means clustering on the matrix obtained from SNMF decomposition, using the centroid of the pure human voice time-frequency spectrum as the cluster center to enhance the robustness of the separation effect. The human voice separation step involves adjusting the original TASNet to a 2D-TASNet, using the features obtained from K-Means clustering as input to train the 2D-TASNet to obtain the target mask, and using this mask to reconstruct the separated human voice.

[0107] In summary, the 2D-TASNet-based human voice denoising method for computer rooms provided by this invention converts the time-domain signal into a two-dimensional time-frequency spectrum through short-time Fourier transform, preserving the joint time-frequency features of the signal and providing a richer information foundation for subsequent processing. By combining sparse non-negative matrix factorization (SNMF) to perform feature decomposition on the amplitude spectrum, and leveraging the sparsity of human voice in the time-frequency domain, non-negativity constraints and L1 sparsity constraints are applied to ensure that the resulting basis matrix and coefficient matrix better reflect the feature differences between human voice and noise, effectively separating representative local features and laying a high-quality feature foundation for subsequent clustering and separation. K-Means clustering is then used to cluster the SNMF-decomposed matrix, further filtering out pure features containing only human voice and eliminating noise-related features. This feature purification process reduces the interference of noise on subsequent network training, allowing 2D-TASNet to focus more on learning the feature patterns of human voice and improving the model's accuracy in recognizing and separating human voice.

[0108] This paper proposes targeted improvements to the native TASNet, adjusting the input to a two-dimensional amplitude spectrum and the output to an ideal proportional mask. The network's internal parameters and loss function are also adapted to better suit the processing requirements of two-dimensional time-frequency features. This improved 2D-TASNet can fully learn the local and global correlations in the time-frequency domain, generating a more accurate target mask, thus effectively separating human voices from mixed signals. Simultaneously, the end-to-end network design simplifies the process, reduces manual intervention, and improves separation efficiency. Given the complex background noise in computer room environments (such as equipment operating noise and electromagnetic interference), this method utilizes the sparsity constraints of SNMF and K-Means clustering to specifically suppress the persistent or sudden noise unique to computer rooms. Combined with the deep mining capabilities of 2D-TASNet for two-dimensional time-frequency features, it maintains high human voice separation quality even in low signal-to-noise ratio computer room environments, enhancing the robustness and practicality of the method in real-world scenarios. By performing a dot product operation between the target mask and the time spectrum of the mixed signal, and combining it with the inverse short-time Fourier transform to reconstruct the human voice, it is possible to suppress noise while preserving the original spectral characteristics and temporal continuity of the human voice to the greatest extent, ensuring that the separated human voice is clear and intelligible, and realizing the effective extraction of human voice signals in computer room scenarios.

Claims

1. A method for human voice noise reduction in a computer room based on 2D-TASNet, characterized in that, Includes the following steps: S1： Preprocessing: Perform a short-time Fourier transform on the mixed time-domain signal of human voice and background noise to convert the one-dimensional time-domain signal into a two-dimensional time-spectrum signal; S2: Feature decomposition: Based on the sparsity of human voice, sparse nonnegative matrix factorization (SNMF) is used to extract features of the input mixed signal; S3: Feature clustering: After the amplitude spectrum is factored by sparse nonnegative matrix factorization (SNMF), the resulting matrix needs to be clustered using K-Means. S4: Voice separation: This is achieved through an end-to-end separation system. It is based on the original TASNet, with the input modified to a two-dimensional amplitude spectrum and the output modified to an ideal proportional mask. The parameters and loss function of the internal network structure are also adjusted to adapt the input and output. The features obtained by K-Means clustering are used as the input of 2D-TASNet to train 2D-TASNet to obtain the target mask. S5: Reconstruct the separated human voice using the target mask.

2. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 1, characterized in that, The specific process of step S1 is as follows: The time-frequency domain signal is obtained by performing a short-time Fourier transform (STFT) through a Hamming window with a window length of 256 points and a window shift of 128 points. s ( f , t ); ； in, It is a window function. f Represents frequency, s ( t () is a mixed signal of human voice and background noise.

3. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 1, characterized in that, The specific process of feature decomposition in step S2 is as follows: S21: Take the modulus value of the two-dimensional time-frequency signal to obtain the amplitude spectrum: V( f , t )=| s ( f , t )|； S22: The input matrix V is decomposed into an r-dimensional basis matrix W and a coefficient matrix H, with constraints including non-negativity of elements and L1 sparsity constraints, which are reflected in the coefficient matrix H. An SNMF objective function with Euclidean distance is then defined. ； ； in, λ It is a regularization parameter used to control sparsity and reconstruction error; the dimension of V is... f * t , W The dimension is f * r , H The dimension is r * t , r For decomposition dimensions.

4. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 3, characterized in that, Execute the iterative process: S23: Initialize W and H as random positive matrices, apply sparsity constraints to H, then project each row of H to be non-negative, and set the L1 criterion to achieve the desired sparsity; S24: Project each column of W as nonnegative: ； S25: Project each row of H to be non-negative, and set the unit L1 specification to achieve the preset sparsity: ； When the cost function converges, exit the loop and output matrices W and H; otherwise, continue iterating.

5. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 1, characterized in that, The specific process of feature clustering in step S3 is as follows: S31: The decomposed basis matrix W is a vector set consisting of K basis vectors. The coefficient matrix H consists of K coefficient vectors, and the vector set W consists of K basis vectors, as shown below: ； ； S32: The basis matrix W and coefficient matrix H contain human voice and noise features distributed across different frequency bands, by... w k and h k The submatrix obtained by multiplication F k : ； submatrix F k These are characteristics of the source signal. w k and h k Clustering is performed to obtain the target features. The clustering results in pure features that contain only human voices and no noise.

6. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 1, characterized in that, The specific process of voice separation in step S4 is as follows: S41: The encoder maps the 2D input features of the amplitude spectrum of the mixed signal to a higher-dimensional space. ； S42: Separation is achieved in high-dimensional space using an estimator; S43: The decoder recovers the high-dimensional vector calculated by the estimator into a mask of the same dimension as the encoder input. 。 7. The method for human voice noise reduction in a computer room based on 2D-TASNet according to claim 1, characterized in that, The specific process of reconstructing the separated human voice using the target mask in step S5 is as follows: After performing a time-spectrum dot product of the mask and the mixed signal, a time-spectrum containing only the target human voice is obtained. Then, an inverse STFT is performed to obtain the separated human voice. The specific formula is as follows: ； Where s1(t) is the separated human voice signal, M is the estimated mask, S is the time spectrum of the mixed signal, and ISTFT is the inverse Fourier transform.

Images