Audio De-noising Analysis Using Diagonal and Non-Diagonal Estimation Techniques

Sugat Pawar, Vaishali Gaderao, Rahul Jadhav
2014 International Journal of Technology and Science   unpublished
Audio signals are often contaminated by background environment noise and buzzing or humming noise from audio equipment. Audio denoising aims at attenuating the noise while retaining the underlying signals. Major audio signal denoising techniques are based on attenuation in time-frequency signal representations. The time-frequency audio denoising algorithms of MMSE-LSA estimator of Diagonal Estimation technique and Adaptive time-frequency block Thresholding of non-diagonal estimation techniques
more » ... imation techniques are considered. Removing noise from audio signals requires a non-diagonal processing of time-frequency coefficients to avoid producing "musical noise". Time-frequency audio denoising procedures compute a Short-Time Fourier Transform (STFT) of the noisy signal and process the resulting coefficients to attenuate the noise. Simple time-frequency denoising algorithms compute each attenuation factor only from the corresponding noisy coefficients and are thus called diagonal estimators.