RESUMO
This paper presents an integrated speech enhancement (SE) method for the noisy MRI environment. We show that the performance of SE system improves considerably when the speech signal dominated by MRI acoustic noise at very low SNR is enhanced in two successive stages using two-channel SE methods followed by a single-channel post processing SE algorithm. Actual MRI noisy speech data are used in our experiments showing the improved performance of the proposed SE method.
Assuntos
Acústica , Imageamento por Ressonância Magnética/instrumentação , Imageamento por Ressonância Magnética/métodos , Som , Fala , Algoritmos , Percepção Auditiva , Desenho de Equipamento , Humanos , Movimento , Ruído , Acústica da Fala , Percepção da FalaRESUMO
Performance of two Adaptive (nLMS and Normalized Sign-error LMS) and a single channel (LogMMSE) speech enhancement algorithms are tested on a floating point DSP to reveal their effectiveness in enhancing speech corrupted in noisy MRI environment with very low SNR. The purpose of experiments is to reduce the fatigue of the listener by eliminating the strong MRI noise. The experiments use actual data set collected from a 3-Tesla MRI machine. Results of the experiments and performance of the speech enhancement system are presented in this paper. The speech enhancement system is automated. Our experiments reveal that after enhancement of the speech signal using Sign-Error LMS, the residual noise shows characteristics of white noise in contrast to the residual noise of the other algorithms which is more structured. It is also shown that the Sign-Error LMS offers fast convergence in comparison to the other two methods.
