Speech recognition system with huge vocabulary

A technology of speech recognition and word recognition, applied in speech recognition, speech analysis, instruments, etc., can solve problems such as limited number of words

Active Publication Date: 2008-12-17
NUANCE COMM INC
View PDF1 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, the main difficulty in increasing the number of recognizable words by state-of-the-art LVCSR is the need to collect a sufficiently large lexicon
Al

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 with huge vocabulary
  • Speech recognition system with huge vocabulary
  • Speech recognition system with huge vocabulary

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] In a structure such as a standard Large Vocabulary Continuous Speech Recognizer (LVCSR), User Dictionary (ULX) and Language Model (LM) are basic components. Together they limit the amount of recognizable words.

[0041] The speech recognition system presented in this paper overcomes this limitation, and we call the speech recognition system presented in this paper the Huge Continuous Speech Recognizer (HVCSR), because it can recognize a huge number of words, and in principle can recognize an unlimited number of words. HVCSR does not have a traditional LM, it uses a so-called large word lexicon (HwLex) instead of a traditional ULX to determine the allowed words of the language actually used. HwLex stores actual language words and their phonetic symbols. HwLex will be described in further detail below. Compared with LVCSR, in HVCSR, the information sources are combined differently, thereby being able to handle a large number of recognizable words. Typically, HwLex is 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 deals with speech recognition, such as a system for recognizing words in continuous speech. A speech recognition system is disclosed which is capable of recognizing a huge number of words, and in principle even an unlimited number of words. The speech recognition system comprises a word recognizer for deriving a best path through a word graph, and wherein words are assigned to the speech based on the best path. The word score being obtained from applying a phonemic language model to each word of the word graph. Moreover, the invention deals with an apparatus and a method for identifying words from a sound block and to computer readable code for implementing the method.

Description

technical field [0001] This invention relates to speech recognition systems for recognizing words from sound blocks, and more particularly to continuous speech recognizers. In addition, the present invention also relates to an apparatus and method for recognizing words based on sound blocks, and computer readable codes for implementing the method. Background technique [0002] In a speech recognition system, an input sound block is processed by a computer system that converts the sound characteristics of the spoken content of the sound block into recognized words. Speech recognition is a complex job involving many steps. The first step usually involves some kind of acoustic feature extraction, where acoustic features representing words or word parts are extracted from sound blocks from acoustic sources. The acoustic features are then scored, with an acoustic score describing the probability that a particular word or word part would produce a certain feature at a given posi...

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): G10L15/08G10L15/18G10L15/187
CPCG10L15/08G10L15/187G10L2015/025G10L15/28
Inventor Z·萨费
Owner NUANCE COMM INC
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap