Abstract | ||
---|---|---|
Speech enhancement algorithm is proposed for various noise environments. It is based on calculating the ratio of the standard deviation (STD) of the noisy power spectrum in time-frequency bins to its normalized time-frequency average. If the ratio greater than the adaptive threshold, speech is considered to be present. Our adaptive algorithm tracks the threshold and controls the trade-off between residual noise and distortion. The algorithm has the advantages of simplicity and light computational load for estimating the noise. Specifically, this algorithm reduces the residual noise significantly, and is superior to conventional methods. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1109/FBIT.2007.10 | FBIT |
Keywords | Field | DocType |
noisy power estimates,time-frequency bins,adaptive threshold,speech intelligibility,speech enhancement algorithm,distortion,adaptive algorithm,standard deviation,masking,time-frequency bin,residual noise,conventional method,various noise environment,light computational load,normalized time-frequency average,speech enhancement,noisy power spectrum,time-frequency analysis,power spectrum,time frequency,time frequency analysis,adaptive thresholding | Speech enhancement,Masking (art),Computer science,Speech recognition,Spectral density,Time–frequency analysis,Adaptive algorithm,Distortion,Standard deviation,Intelligibility (communication) | Conference |
ISBN | Citations | PageRank |
978-0-7695-2999-8 | 0 | 0.34 |
References | Authors | |
2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Soojeong Lee | 1 | 51 | 5.93 |
Soon-Hyob Kim | 2 | 1 | 2.07 |