The study of the characteristics of LSP parameters is reported,and an efficient coding method for LSP parameters of Chinese speech is proposed by using the sensitivity of LSP parameters to the spectral envelope .

Phoneme segmentation plays a key role in the recognition of large vocabulary continuous Chinese speech recognition.

Chinese speech waveform coding synthesis is based on the time domain pitch synchronous overlap add(PSOLA)method and a Chinese syllable dictionary.

The most direct influence on foreign brand word is from modern Chinese pronunciation and character.

in primary classes,investigate and record their Chinese pronunciation .

It is regarded as a main objective of ethnic students\'learning Chinese in Xinjiang to acquire standard Chinese pronunciation.

Entropy and fossilization phenomena are two biggest problems which have bothered both students and teachers for a long time in learning Chinese phonetics.

The difficulty in the modern Chinese phonetics teaching should be chosen.

There exists faults between the phonetics of the ancient Chinese and that of the modern standard mandarin in teaching Chinese phonetics in the teachers universities.

Thus there are errors in pronouncing English fricatives.

Application of the book number mixed with the Chinese phonetic alphabets and the four-corner system numerals for the authors names;

Mixed book number of the Chinese phonetic alphabets and the four-corner system numerals for the author s name;