Abstract:In this paper our approach to the lexical tone recognition of Chinese continuous speech is presented. The Mixed Gaussian Continuous Probability Model (MGCPM) (1) is used for the tone modeling, and the quadric curve is adopted to simulate the Fundamental frequency (F0) contour, whose three coefficients are calculated and taken as the features of the tone models. The tone variety in continuous Chinese speech recognition is an issue that must be faced in the tone modeling. There are two kinds of tone varieties, the change from canonical one to non- canonical one without changing the pitch trend and that from one to another different one. In order to reduce the negative influence caused by the tone varieties, an iterative method is proposed to distinguish the syllables which have tone varieties and remove them from the whole training data, and then the Tone Variety Matrix (TVM) is introduced for improving the performance of tone models. Experiments have been done based on the continuous Chinese speech database named "863" database. The top1 and top2 accuracy for baseline MGCPM is 67% and 90%, while that for MGCPM incorporated with TVM is 70% and 92%.

Recognition of Chinese Continuous Speech Based on 3-Dimension Viterbi Search

Study on the Integration of Phonetic and Prosodic Probability Based on 3-Dimension Viterbi Search

Continuous Speech Recognition Based on the Triphone DDBHMM

TONE RECOGNITION OF CHINESE CONTINUOUS SPEECH

Pronunciation recognition and assessment for mandarin Chinese

Algorithm for Mandarin Continuous Speech Recognition Based on Context-Dependent Unit Between Syllables

Large Vocabulary Mandarin Chinese Continuous Speech Recognition System Based on Tonal Triphone

Feature Selection In Mandarin Large Vocabulary Continuous Speech Recognition

ON CHINESE CONTINUOUS SPEECH RECOGNITION SYSTEM AND KNOWLEDGE BASED SEARCH STRATEGIES SONG Zhanjiang

Mandarin Continuous Digit Speech Recognition System

Chinese Vowels Recognition Method Based on Non-Homogeneous Hidden Markov Model

Research on Continuous Chinese Speech Segmentation Based on Pitch Periodic Track

Research on Segmentation of Continuous Chinese Speech

Research on continuous Chinese speech recognition system and knowledge based search strategies

Chinese Continuous Speech Endpoint Detection Method Based on Vowel Detection

An Efficient Computation Algorithm In Mandarin Continuous Speech Recognition

Mandarin Name Recognition Based on Variable Frame Rate HMM Training

Research on Inter-Syllable Context-Dependent Acoustic Unit for Mandarin Continuous Speech Recognition.

Detecting Tone Errors In Continuous Mandarin Speech

Towards Large Vocabulary Mandarin Chinese Speech Recognition

A Chinese speech recognition system