A New Method Based on Spectral Subtraction for Speech Dereverberation

By K. Lebart, J. Boucher, P. Denbigh. 2001

325 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

The NTT CHiME-3 system: Advances in speech enhancement and recognition for mobile multi-microphone devices (2015, methodology): “An important feature of this method is that, unlike approaches based on spectral subtraction [25, 26], dereverberation is carried out with a linear time-invariant ﬁ lter and thus introduces little artefact.”
Noise Model Transfer: Novel Approach to Robustness Against Nonstationary Noise (2013, methodology): “These observations motivated us to apply NMT to reverberant speech recognition, whereby we will be able to exploit existing reverberation power spectrum estimation techniques and the widely used VTS enhancement method.”
Speech Enhancement of Noisy and Reverberant Speech for Text-to-Speech (2018, methodology): “This system used a RIR-modeling dereverberation method [13] based on the hypothesis that a RIR can be represented as a Gaussian stationary”
Learning Spectral Mapping for Speech Dereverberation and Denoising (2015, background): “Other studies dealt with dereverberation by exploiting the properties of speech such as modulation spectrum [2], power spectrum [21], and harmonic structure [40], [30].”