Signal processing method and device, electronic equipment and storage medium

Abstract

The invention provides a signal processing method and device, electronic equipment and a storage medium. According to the specific implementation scheme, the method comprises the following steps: acquiring an original signal collected by a microphone, carrying out short-time Fourier transform on the original signal collected by the microphone to obtain a frequency domain signal, estimating a reverberation spectrum of the frequency domain signal by adopting a Kalman filter, carrying out reverberation suppression on the frequency domain signal according to the reverberation spectrum to obtain a reverberation-removed signal, subjecting the spectrum parts in the reverberation-removed signal to cepstrum smoothing through corresponding smoothing coefficients, and performing short-time inverse Fourier transform according to the smoothed reverberation-removed signal to obtain a target signal. The reverberation spectrum is adaptively estimated through the Kalman filter, and the estimated reverberation spectrum is suppressed through a spectrum enhancement technology to obtain the reverberation-removed signal, so that the frequency spectrum features are enhanced, and the voice quality is improved; and meanwhile, the music noise problem introduced in the spectrum enhancement process is solved through the cepstrum smoothing method, and the voice quality is further improved.

Description

technical field [0001] The present application relates to the technical field of signal processing, and in particular to signal processing methods, devices, electronic equipment and storage media. Background technique [0002] Reverberation is a type of acoustic noise that appears in an enclosed space through multiple reflections and diffractions of the sound from the walls and objects in the room. In a reverberant environment, the previously emitted speech signal is reflected and overlapped and masked by the current speech signal, which leads to blurring of spectral features, thereby affecting speech intelligibility and speech quality. Among them, the signal with reverberation is composed of three consecutive parts: direct sound, early reflections and late reverberation. Studies have shown that early reflections can help improve human perception, while late reverberation can produce overlapping-masking effects on direct sounds, affecting speech quality and intelligibility,...

AI Technical Summary

Problems solved by technology

[0003] At the same time, in the statistical processing of noise signals, estimation errors are unavoidable, so outliers will appear in the process of spectral gain adaptation, resulting in an auditory phenomenon called musical noise

Images