Unlock instant, AI-driven research and patent intelligence for your innovation.

Robust speech enhancement method and system based on mouth-binaural room impulse response

An impulse response, binaural room technology, applied in speech analysis, instruments, etc., can solve the problem of not considering the influence of the acoustic model of the binaural near-field model, so as to solve the azimuth estimation error and the position disturbance is too sensitive, and suppress the far-field synchronization. To interfere with the sound source, the effect of high robustness

Active Publication Date: 2021-12-07
INST OF ACOUSTICS CHINESE ACAD OF SCI
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method is only applicable to free-field acoustic models, and does not consider the influence of scatterers such as the human head and torso in the binaural near-field model on the acoustic model.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Robust speech enhancement method and system based on mouth-binaural room impulse response
  • Robust speech enhancement method and system based on mouth-binaural room impulse response
  • Robust speech enhancement method and system based on mouth-binaural room impulse response

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0107] Figure 1 and figure 2 As shown, Embodiment 1 of the present invention provides a robust speech enhancement method based on mouth-binaural room impulse response. The Brüel&Kjaer 4128C artificial head and torso simulator (Head And Torso Simulator, HATS) and three MEMS dual microphone arrays with an aperture of about 1.5 cm were used to measure the multi-channel port-binaural room impulse response. Two dual MEMS microphone chips are fixed on the outside of the wireless earphone, and the wireless earphone is worn in the artificial ear. The front and rear side microphones of the right ear of the artificial head are respectively No. 1 microphone and No. 2 microphone, and the front and rear side microphones of the left ear of the artificial head are No. 3 microphone and No. 4 microphone respectively. Another double MEMS microphone chip is fixed at the front end bracket of the artificial mouth, and the left and right sides of the chip are No. 5 microphone and No. 6 microphone....

Embodiment 2

[0115]Embodiment 2 of the present invention provides a robust speech enhancement system based on mouth-binaural room impulse response. The system includes: a signal acquisition module, a weight vector calculation module and an enhanced output module;

[0116] The signal acquisition module is used to obtain the original signal with noise;

[0117] The weight vector calculation module is used to extract several groups of multi-channel port-binaural room impulse responses from the pre-established multi-channel port-binaural room impulse response database; The transformation is converted into a frequency-domain transfer function, and the frequency-domain transfer function is used as a steering vector to form a steering vector matrix; the eigenvalue decomposition of the steering vector matrix is ​​performed, and the main eigenvectors are constrained while minimizing the output signal energy. By solving the convex The optimization problem is calculated to obtain the beamforming wei...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a robust speech enhancement method and system based on mouth-binaural room impulse response. The method comprises the following steps: acquiring an original noisy signal; extracting a plurality of groups of multi-channel port-binaural room impulse responses from a pre-established multi-channel port-binaural room impulse response database; converting the multi-channel port-binaural room impulse response into a frequency domain transfer function through Fourier transform, and taking the frequency domain transfer function as a steering vector to form a steering vector matrix; carrying out eigenvalue decomposition on the steering vector matrix, constraining main eigenvectors while output signal energy is minimized, and calculating a beam forming weight vector by solving a convex optimization problem; and performing weighted summation on the original noisy signal by using the beam forming weight vector, and outputting an enhanced voice signal. The method provided by the invention can effectively suppress a far-field co-directional interference sound source, and has higher robustness compared with a traditional near-field beam former.

Description

technical field [0001] The invention relates to the field of speech enhancement. In particular, it relates to a robust speech enhancement method and system based on mouth-binaural room impulse response. Background technique [0002] Nowadays, earphones have received more and more people's favor and attention, especially true wireless earphones. In the past two years, both in the academic field and in the commercial field, they have been one of the products that experts and scholars have focused on research, development and mass production. With the rapid development of technologies such as Bluetooth, audio coding, integrated circuits, and artificial intelligence, people rely more and more on headset devices, and their requirements for headset communication quality are also getting higher and higher. However, in the practical application of earphone communication, the complex noise environment will lead to serious degradation of communication quality. Therefore, the voice e...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L21/0224G10L21/0232
CPCG10L21/0232G10L21/0224
Inventor 柯雨璇侯畅郑成诗李晓东
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI