Абстрактный

COMPARATIVE STUDY OF MFCC AND LPC FOR MARATHI ISOLATED WORD RECOGNITION SYSTEM

Leena R Mehta , S.P.Mahajan , Amol S Dabhade

This Paper presents Marathi database and isolated word recognition system using Mel-frequency cepstrum coefficients (MFCC) and vector quantization (VQ) technique. It also compares the performances of MFCC and LPC features under VQ environment. Marathi speech database is recorded in noisy environment aiming language learning tool as an application. The database consists of simple Marathi words starting with both vowels and consonants. Each word has been repeated 10 times by one male and one female speaker. This paper presents comparative plots of MFCC and LPC features.