Adaptive Echo Noise Elimination for Speech Enhancement of Tamil letter ‘Zha’

Acoustic echo depends on time delay between initial and reflected sound wave, strength of reflected sound. In the speech processing of letter ‘zha’ [11], echo of the recorded voice gives the spurious results. Such complexity can be avoided by suitable pyramidal method like adaptive filtering technique. Adaptive filtering tries to adjust these parameters with the aim of meeting some well-defined target, which depends upon the state of the system and surroundings. In speech recognition, the acoustic echo gives the faulty results. Objective of this paper is to analyze the performance of various adaptive filtering algorithms for acoustic echo cancellation in recorded speech enhancement of the letter ‘Zha’ in Tamil language. These algorithms are simulated in MATLAB and compared with the performance of those algorithms based on parameters such us computational complexity, convergence rate and amount of echo attenuation.