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

Voice recognition method based on layered circulation neural network language model

A cyclic neural network and language model technology, applied in biological neural network models, speech recognition, neural learning methods, etc., can solve the problems of low recognition accuracy and large storage space.

Inactive Publication Date: 2017-05-31
SHENZHEN WEITESHI TECH
View PDF0 Cites 23 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problems of low recognition accuracy and large storage space, the purpose of the present invention is to provide a speech recognition method based on a layered recurrent neural network language model. First, the character-level language modeling of RNN is used, and then an external clock and Reset signal extended RNN structure, character-level language modeling with hierarchical RNN, and finally 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
  • Voice recognition method based on layered circulation neural network language model
  • Voice recognition method based on layered circulation neural network language model
  • Voice recognition method based on layered circulation neural network language model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0048] It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0049] figure 1 It is a system flowchart of a speech recognition method based on a layered recurrent neural network language model of the present invention. Mainly including character-level language modeling using RNN, extending RNN structure with external clock and reset signal, character-level language modeling with hierarchical RNN and performing speech recognition.

[0050] Among them, the described extended RNN structure with external clock and reset signals, most types of RNNs can be generalized as

[0051] the s t =f(x t ,s t-1 ) (1)

[0052] the y t =g(s t ) (2)

[0053] where x t is the input, s t is the state, y t is the output at time step t, ...

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 voice recognition method based on a layered circulation neural network language model. The method mainly comprises steps of character-level language modeling using RNN, expansion of an RNN structure by use of an external clock and a reset signal, character-level language modeling with graded RNN and voice recognition. According to the invention, the traditional single-clock RNN character-level language model is replaced by layered circulation neural network-based language model, so quite high recognition precision is achieved; quantity of parameters is reduced; vocabulary of the language model is huge; required storage space is quite small; and a layered language model can be expanded to process information of quite long period, such as sentences, topics or other contexts.

Description

technical field [0001] The invention relates to the field of speech recognition, in particular to a speech recognition method based on a layered recurrent neural network language model. Background technique [0002] With the development of modern technology, character-level language models (CLMs) based on recurrent neural networks (RNN) are widely used in speech recognition, text generation, and machine translation. It is useful for modeling words not seen in nature. However, their performance is usually much worse than word-level language models (WLMs). Moreover, statistical language models require a large storage space, usually exceeding 1GB, because not only a large number of vocabularies, but also their combinations need to be considered. [0003] The present invention proposes a speech recognition method based on a hierarchical recurrent neural network language model, and its hierarchical RNN architecture is composed of multiple modules with different clock rates. De...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G10L15/16G10L15/183G06N3/08
CPCG06N3/084G10L15/16G10L15/183
Inventor 夏春秋
Owner SHENZHEN WEITESHI TECH
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