Audio data processing method, audio data processing device and storage medium

Abstract

The embodiment of the invention discloses an audio data processing method, an audio data processing device and a storage medium, wherein the method is applied to the field of artificial intelligence and comprises the steps of obtaining multipath audio data in an environment space; obtaining a voice data set on the basis of the multipath audio data; respectively generating enhanced voice information corresponding to the voice data set in a plurality of enhancing directions; matching voice hidden features in the enhanced voice information with target matching words; determining the enhancing direction corresponding to the enhanced voice information with the highest matching degree with the target matching words as the target audio direction; obtaining voice frequency spectrum features in theenhanced voice information; obtaining the voice frequency spectrum features in the target audio direction in the voice frequency spectrum features; and performing voice verification on the voice hidden features and the voice frequency spectrum features in the target voice direction on the basis of the target matching words to obtain a target verification result. By using the method, the device and the storage medium provided by the invention, the voice control accuracy can be improved, and the misrecognition rate is effectively reduced.

Description

Technical field [0001] The present invention relates to the field of Internet technology, in particular to an audio data processing method, device and storage medium. Background technique [0002] With the development of science and technology, smart voice devices are gradually being popularized and used. In the implementation of smart voice devices, the voice interaction between voice commands and smart voice devices is an important technical means to realize device intelligence. In the current voice interaction system of intelligent voice equipment, during the process of picking up the wake-up and command words, it will inevitably be interfered by the noise of the external environment and other people’s speech. If the interference noise is too strong, it will seriously affect the voice interaction system. Sensitivity, in turn, causes the wake-up failure or false wake-up phenomenon during the device wake-up process, that is, the accuracy of the device wake-up cannot be ensured. ...

AI Technical Summary

Problems solved by technology

In the current voice interaction system of intelligent voice equipment, in the process of picking up wake-up words and command words, it will inevitably be disturbed by external environmental noise and other people's speech. If the interference noise is too strong, it will seriously affect the voice interaction system. Sensitivity, so that there will be wake-up failure or false wake-up during the wake-up process of the device, that is, the accuracy of the wake-up of the device cannot be guaranteed

Images