Signal processing method, device and storage medium

Abstract

Embodiments of the present disclosure provide a signal processing method, device, and storage medium. The signal processing method includes: obtaining an audio signal to be processed; obtaining a frequency domain signal corresponding to the audio signal to be processed by performing time-frequency transformation on the audio signal to be processed, and calculating energy values ​​of multiple frequency points in the frequency domain signal, and then obtaining A preset mapping relationship, the mapping relationship includes the frequency response of each frequency point under different energy values, and according to the energy value of the frequency point and the above mapping relationship, the normalized energy value corresponding to the frequency point is obtained. The embodiment of the present disclosure determines the normalized energy value corresponding to the frequency point based on the two-dimensional mapping relationship including the frequency point and the energy value, which can improve the distributed voice wake-up decision using the normalized energy value accuracy.

Description

technical field [0001] The present disclosure relates to signal processing technologies, and in particular, to a signal processing method, device and storage medium. Background technique [0002] With the rapid development of the Internet of Things, smart homes are interconnected with multiple devices. In this case, distributed voice wake-up decisions are the first step in responding to user needs. However, due to the design differences of various equipment manufacturers in microphone selection and pickup acoustic structure, and even the same equipment has multiple microphone suppliers, the parameters such as noise floor and sensitivity of microphones are quite different. In addition, it is difficult for the microphone to satisfy the theory that the frequency response curve increases linearly with the energy value. Coupled with the subtle differences in the acoustic structure, the relationship between the frequency response curve and the energy value is nonlinear to a certa...

AI Technical Summary

Problems solved by technology

However, due to the design differences of various equipment manufacturers in microphone selection and pickup acoustic structure, even the same equipment has multiple microphone suppliers, so the parameters such as noise floor and sensitivity of microphones are quite different.

In addition, it is difficult for the microphone to satisfy the theory that the frequency response curve increases linearly with the energy value. Coupled with the subtle differences in the acoustic structure, the relationship between the frequency response curve and the energy value is nonlinear to a certain extent, which leads to the use of energy value related parameters. Decreased accuracy of distributed voice wake-up decisions

Images