Voice wake-up detection method, device and equipment and storage medium

Abstract

The invention provides a voice wake-up detection method and device, equipment and a storage medium. The method comprises steps of inputting an audio data frame in a preset range near a target frame inmultiple frames of audio data and a target frame into an acoustic model assembly together, wherein the acoustic model assembly is a feedforward sequence memory neural network model assembly, and theoutput of the acoustic model assembly is a state recognition result of the target frame and at least one frame of audio data in audio data frames within a predetermined range; taking a single frame ofaudio data which is located behind the target frame and does not be processed in the multiple frames of audio data as a next target frame, and iteratively using the acoustic model assembly to processa plurality of subsequent target frames; and comparing the state identification result of the audio data of the plurality of frames in the plurality of frames of audio data with a preset wake-up wordto identify whether the plurality of frames of audio data are wake-up instructions or not. Therefore, good awakening performance can be guaranteed while occupation of equipment end resources is reduced, and the requirement for real-time performance required by awakening is met.

Description

technical field [0001] The present disclosure relates to the field of voice technology, and in particular, to a voice wake-up detection method, device, equipment and storage medium. Background technique [0002] Voice wake-up means that when the user speaks a specific voice command (that is, a wake-up word), the device switches from the sleep state to the working state. The purpose of the wake-up technology is that the user can use voice to operate the device completely without the help of hands; at the same time, using the wake-up mechanism, the device does not need to be in the working state all the time, which greatly saves energy consumption. At present, voice wake-up technology is widely used in various voice-controlled products, such as robots, mobile phones, wearable devices, smart homes, and vehicles. [0003] Generally speaking, such products need to support working in both networked and non-networked environments, and as the first step of interaction, wake-up must...

AI Technical Summary

Problems solved by technology

However, the computing resources on the device side are usually very limited. Whether it is the number of CPU cores, memory size or core frequency, they are far smaller than the computers we usually use, let alone cloud servers.

Images