PERCEPTUAL LINEAR PREDICTION (PLP)

Clean Speech (25.3 dB SNR) Noisy Speech (2.1 dB SNR)