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

Chinese speech identification method with dialect background

A speech recognition and dialect technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of huge database workload, difficulty in sharing and compatibility of speech recognizer resources, inability to take into account commonalities, etc., to achieve convenient use and maintenance, saving workload, performance improvement

Inactive Publication Date: 2004-11-24
BEIJING D EAR TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The disadvantages of this method are: (1) The workload of collecting the database with dialect background is very huge. For so many dialects in Chinese, the collection of database is a huge project
(2) This method cannot take into account the commonality between standard Mandarin and Mandarin with pronunciation background. It only solves the problem through a data-driven method, which is equivalent to completely rebuilding a speech recognizer and giving speech recognizers with different dialect backgrounds Difficulties in resource sharing and compatibility between

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
  • Chinese speech identification method with dialect background
  • Chinese speech identification method with dialect background
  • Chinese speech identification method with dialect background

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] The Chinese speech recognition method with dialect background that the present invention proposes is described in detail as follows in conjunction with embodiment and accompanying drawing:

[0055] The present invention proposes a Chinese speech recognition method with a dialect background, comprising a speech recognizer of standard Chinese Mandarin; it is characterized in that the method comprises the following steps:

[0056] 1) According to the characteristics of a specific dialect, construct a syllable mapping table from Mandarin pronunciation to dialect pronunciation;

[0057] 2) According to the syllable mapping table, expand the search tree in the existing standard Mandarin speech recognizer;

[0058] 3) Replace the search tree in the standard Mandarin speech recognizer with an expanded search tree.

[0059] Above-mentioned steps 1) construct the method embodiment of syllable mapping table, as Figure 5 shown, including the following steps:

[0060] (1) Summar...

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 present invention belongs to the field of computer artificial intelligent and mode identification technology, relates to a Chinese phonetic recognition method with dialect background. Said invention includes: according to the characteristics of specific dialect constructing syllable mapping talbet from common speech pronunciation to dialect pronunciation; according to the syllable mapping table expanding the search tree in the existent standard common speech phonetic recognizer, then replacing the search tree in standard common speech phonetic recognizer with the expanded search tree. Said invention not only can identify standard common speech, but also can identify the common speech with various dialects only by using syllable mapping table.

Description

technical field [0001] The invention belongs to the technical field of computer artificial intelligence and pattern recognition, and in particular relates to a method for recognizing human speech by a computer. Background technique [0002] "Large Vocabulary Continuous Speech Recognition" (LVCSR, referred to as "Speech Recognition") is to use the computer to identify which words correspond to a certain segment of speech based on the language information contained in the continuous sound signal of a person. process. "Large Vocabulary Continuous Speech Recognizer" (Large Vocabulary Continuous Speech Recognizer, referred to as "Speech Recognizer") refers to equipment or software used for speech recognition. Speech recognition is a conversion process from sound signals to text, and can be widely used in almost all aspects including telecommunications, banking, finance, tourism and transportation, public utilities, entertainment, public consumption behavior, and business operati...

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 Patents(China)
IPC IPC(8): G10L15/187G10L15/22
Inventor 郑方
Owner BEIJING D EAR TECH