Unlock instant, AI-driven research and patent intelligence for your innovation.

Classifier training method and device, speech recognition method and device, equipment and storage medium

A training method and technology of speech recognition, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of reducing the accuracy of speech recognition and poor performance of children's voice recognition, and achieve the effect of improving the accuracy and improving the recognition effect.

Pending Publication Date: 2022-04-01
SOUNDAI TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the embodiments of the present invention is to provide a classifier training method, speech recognition method, device, equipment and storage medium to solve the technical problem that the recognition effect of children's voice is not good, thereby reducing the accuracy of speech recognition

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Classifier training method and device, speech recognition method and device, equipment and storage medium
  • Classifier training method and device, speech recognition method and device, equipment and storage medium
  • Classifier training method and device, speech recognition method and device, equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] The technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.

[0035] see figure 1 , figure 1 It is a schematic diagram of the network layer of the classifier to be trained in the embodiment of the present invention. The classifier training method provided by the embodiment of the present invention is applied to figure 1 The classifier shown.

[0036] Such as figure 1 As shown, the classifier may be a classifier based on a deep neural network (Deep Neural Networks, DNN). The classifier includes 1 input layer, 3 hidden layers, 1 feature layer and 1 output layer, where the input layer is used to store the audio feature data of the training samples; the output layer is used to output the classification results; the hidden layer is used to Multi-dimensional expansion of the audio features of the training samples; the feature layer is used to process the audio feat...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a classifier training method, a speech recognition method and device, equipment and a storage medium. The voice recognition method comprises the following steps: acquiring a to-be-recognized audio; preprocessing the to-be-recognized audio to obtain an audio sample corresponding to the to-be-recognized audio; inputting the audio sample into a target classifier to obtain a classification result corresponding to the to-be-recognized audio; and under the condition that the classification result is used for representing that the to-be-recognized audio is the target audio, decoding the to-be-recognized audio to generate a voice recognition result. According to the invention, the to-be-recognized audio is classified through the target classifier, and whether the to-be-recognized audio is the child sound is determined, so that the recognition effect of the child sound is improved; furthermore, under the condition that the to-be-recognized audio is the child voice, decoding processing is performed to generate a voice recognition result, so that the accuracy of voice recognition is improved.

Description

technical field [0001] The present invention relates to the technical field of audio processing, in particular to a classifier training method, a speech recognition method, a device, a device and a storage medium. Background technique [0002] With the wide application of voice interaction technology, more and more electronic devices support the voice recognition function. In the current voice interaction technology, adult corpus that is easy to collect is usually used as audio data. [0003] However, compared with adult audio, children's audio has the characteristics of short vocal tract length, large changes in pronunciation speed, and more meaningless modal particles, and the timbre of children's audio is different from that of adult audio, which leads to the existing Speech recognition methods are not effective in recognizing children's voices, thereby reducing the accuracy of speech recognition. Contents of the invention [0004] The purpose of the embodiments of th...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L15/16G10L15/06G10L25/18
Inventor 陈孝良李良斌
Owner SOUNDAI TECH CO LTD