Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Speech recognition system and method based on transfer neural network acoustic model

A neural network model and neural network technology are applied in the field of speech recognition systems based on the migration neural network acoustic model, which can solve problems such as performance loss and irreversible changes.

Active Publication Date: 2019-07-30
INST OF ACOUSTICS CHINESE ACAD OF SCI +1
View PDF11 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the problem of re-optimizing the model with data in the target language in the prior art, changing the model parameters of the original robust acoustic model irreversibly, and having a certain performance loss in the application of the original robust acoustic model , propose a speech recognition system and method based on transfer neural network acoustic model

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
  • Speech recognition system and method based on transfer neural network acoustic model
  • Speech recognition system and method based on transfer neural network acoustic model
  • Speech recognition system and method based on transfer neural network acoustic model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0062] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0063] An embodiment of the present invention provides a speech recognition system based on a migration neural network acoustic model. The above-mentioned migration neural network refers to migrating model parameters based on a robust neural network acoustic model, and migrating its robust model parameters and model structure to a On the neural network corresponding to the new task, refer to figure 1 shown, including:

[0064] Sig...

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 relates to a speech recognition system and method based on a transfer neural network acoustic model. The system includes a signal processing and feature extraction module, a language model, a decoder and a migrating neural network acoustic model, wherein the migrating neural network acoustic model includes a robust neural network and a directional neural network. According to the system, the model parameters of a robust neural network model is fixed, meanwhile through layer-to-layer transverse connection of the neural networks, information of a robust acoustic model is transferred into a target acoustic model, the performance of an original robust acoustic model is retained, and meanwhile target language is specifically optimized. The problem of fast construction of the robust acoustic model of less-resource languages is solved, and through a mode in which the acoustic model of a language with sufficient data is used for model parameter migration, the performance of the acoustic model of a target less-resource language is improved and the convergence speed of training is increased.

Description

technical field [0001] The invention relates to the field of speech recognition, in particular to a speech recognition system and method based on a transfer neural network acoustic model. Background technique [0002] At present, the automatic speech recognition technology is very mature. Under the technical research of some speech recognition institutions, the recognition accuracy of the automatic speech recognition system can reach 94.5%, which can be said to have reached the human auditory perception ability. But this excellent automatic speech recognition system is limited to a few widely used languages, such as English and French. There are more than 5,000 languages ​​spoken by people all over the world, but only 10 of these 5,000 languages ​​are widely used, they are: Chinese, English, Russian, Spanish, Hindi, Arabic , Portuguese, Bengali, German and Japanese. For other languages, due to the small number of users, it is difficult to collect their speech data, and the...

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): G10L15/02
CPCG10L15/02G10L15/063G10L25/30
Inventor 张鹏远刘丹阳徐及颜永红
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com