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

Spoken language voice recognition method based on statistic model and grammar rules

A statistical model and speech recognition technology, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as the large number of rules, the language model is difficult to adapt to spoken language description, and it is difficult to consider new sentences.

Inactive Publication Date: 2009-12-16
北京森博克智能科技有限公司
View PDF0 Cites 37 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0018] The above language model is difficult to adapt to the description of colloquial language mixed with mantras, ambiguous pronunciation, etc.
In particular, these phenomena related to colloquial languages ​​cannot be predicted, so it seems difficult to rely on their own properties to design grammars based on grammatical rules
[0019] Furthermore, the number of rules required for overlay application is large, and it is difficult to consider new sentences to be added to the dialogue without modifying such existing rules

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
  • Spoken language voice recognition method based on statistic model and grammar rules
  • Spoken language voice recognition method based on statistic model and grammar rules
  • Spoken language voice recognition method based on statistic model and grammar rules

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0070] The present invention will be further described below in conjunction with the accompanying drawings.

[0071] figure 1 It is a schematic diagram of the framework of the present invention, Figure 6 It is a schematic flow chart of the system of the present invention, such as figure 1 and Figure 6 As shown, the system is mainly composed of four parts: S1-acoustic model training, S2-language model training, S3-front-end processing, and S4-recognition and decoding. The system flow is as follows:

[0072] The S1-acoustic model training part of the process is as follows:

[0073] 1. S1-1, feature extraction. According to the frame length of 25 milliseconds and the frame shift of 10 milliseconds, the 12-dimensional MFCC features are extracted, and the 1-dimensional energy features are added to form a total of 13-dimensional static features. The dynamic features take the first-order and second-order difference features to obtain a 39-dimensional acoustic featu...

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 a voice recognition method based on statistic language model, combining grammar rules, and oriented to language speaking recognition application. The invention comprises an acoustic model training, a language model training, a front-end processing, and a recognizing decoding. Based on the N element grammar statistic model, and grammar rule network, the language model of the invention is capable of effectively processing phenomenon such as out-of-vocabulary words, slogan, blurred pronunciation, fast switching between sentences, so that higher recognition rate is achieved with the proviso that naturalness of voice recognition is ensured.

Description

technical field [0001] The invention relates to the technical field of automatic speech recognition, and is a speech recognition method based on a statistical language model, combined with grammatical rules, and oriented to spoken language recognition applications. Background technique [0002] Information systems or control systems are increasingly using voice interfaces to interact quickly and directly with users. As the functions of these systems are becoming more and more complex and the required dialogue methods are becoming richer, people are Enter the field of large-vocabulary spoken continuous speech recognition. [0003] The design of a continuous speech recognition system with a large vocabulary requires the generation of a language model that determines the probabilities of a sequence of words. [0004] For spoken language recognition, this language model must reproduce the speech patterns typically used by users of the system: repetitions, irrelevant parentheses...

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): G10L15/06G10L15/14G10L15/18G10L15/187
Inventor 王辉
Owner 北京森博克智能科技有限公司
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