Phoneme based voice recognition method and system

Abstract

A voice recognizing method and system based on the phoneme includes: A) the analog voice signal is transferred into digital voice signal; B) the short-time zero-crossing ratio is detected, if the short-time zero-crossing ratio is less than the preset value, it is judged to sonant to processed as sonant, if the short-time zero-crossing ratio is higher than the preset value, it is judged to surd to processed as surd; C) the data after pretreatment is spectrum transformed to pick up character; D) the character data is analyzed; E) the phoneme sequence is output according to the analyzed result. The voice recognizing method and system can introduce different process method to surd and sonant, specially the sonant phoneme is modeled based on the single keynote cycle spectrum; it resolves the defect of current voice input recognizing system. It possesses advantages of high recognizing efficiency, high accuracy and high stability.

Description

technical field [0001] The invention relates to the technical field of computer speech recognition, in particular to a phoneme-based speech recognition method and system. Background technique [0002] The fast Fourier transform-FFT of sequences is one of the most important tools for discrete-time signal analysis and processing. If the signal is a sequence of finite length, the frequency spectrum of the sequence can be obtained by directly performing FFT operation on the sequence. For analog signals, when using FFT for spectrum analysis, the signal must first be sampled to make it a discrete signal. According to the sampling theorem, the sampling frequency fs should be greater than twice the highest frequency of the signal. According to the relationship between digital frequency and analog frequency, the analog frequency resolution can be obtained when using N-point FFT for spectrum analysis: [0003] ΔF=fs / N -----------------------------------------(1) [0004] Therefore...

AI Technical Summary

Problems solved by technology

The continuous speech recognition system based on this technology has the following disadvantages: 1. The speaker's pronunciation is required to be very standard; 2. The recognition environment is required to be very quiet and the environmental noise is small; Recognition; 4. The recognition rate is related to the topic, that is, it is related to the content of the template library, and words that are not in the template cannot be recognized; 5. The recognition system is required to establish a large-capacity recognition template library. 6. The repeatability is very poor, and the same In a word, if it cannot be recognized correctly, the recognition results will be different multiple times

[0013] The existing continuous speech recognition system adopts mixed recognition of unvoiced and voiced sounds, samples the speech signal with a fixed duration, and extracts speech recognition features. Because it cannot guarantee that each sampling is for a single phoneme, the extracted recognition features have poor stability and poor recognition effect. far from ideal

Images