Deep learning speech extraction and noise reduction method fusing bone vibration sensor and microphone signals

Abstract

The invention relates to a deep learning noise reduction method fusing bone vibration sensor and microphone signals. The deep learning noise reduction method comprises the following steps: enabling abone vibration sensor and a microphone to collect audio signals, and respectively obtaining a bone vibration sensor audio signal and a microphone audio signal; inputting an audio signal of the bone vibration sensor into a high-pass filtering module, and carrying out high-pass filtering; inputting the bone vibration sensor audio signal subjected to high-pass filtering or the signal subjected to frequency band widening and a microphone audio signal into a deep neural network module; and obtaining the voice after noise reduction through prediction by the deep neural network module. According to the deep learning noise reduction method fusing bone vibration sensor and microphone signals, signals of a bone vibration sensor and a traditional microphone are combined, and the strong modeling capability of the deep neural network is utilized to realize high human voice restoration degree and extremely strong noise suppression capability, and the problem of human voice extraction in a complex noise scene can be solved, and the target human voice is extracted, and the interference noise is reduced, and a single-microphone structure can be adopted to reduce the cost. In addition, the signals obtained after the frequency band of the bone vibration sensor audio signals is broadened can be directly used as output.

Description

technical field [0001] The invention relates to the technical field of voice noise reduction for electronic equipment, and more specifically, to a deep learning noise reduction method for fusing bone vibration sensors and microphone signals. Background technique [0002] Speech noise reduction technology refers to the separation of speech signals from noisy speech signals. This technology has a wide range of applications, usually with single-microphone noise reduction technology and multi-microphone noise reduction technology. However, there are some defects in traditional noise reduction technologies. Traditional The single-microphone noise reduction technology presupposes that the noise is stationary noise, which has low adaptability and large limitations; while the traditional multi-microphone noise reduction technology requires two or more microphones, which increases the cost. Higher requirements limit the structural design of the product. Moreover, the multi-microphone...

AI Technical Summary

Problems solved by technology

[0012] The technical problem to be solved by the present invention is how to use a deep learning noise reduction method that fuses bone vibration sensors and microphone signals to solve the problem of multi-microphone restrictions on product structure, high cost, and traditional single-microphone noise reduction in the prior art. Technical limitations and other issues

Images