We introduce a new speaker independent method for reducing wind noise in single-channel recordings of noisy speech. The method is based on non-negative sparse coding and relies on a wind noise dictionary which is estimated from an isolated noise recording. We estimate the parameters of the model and discuss their sensitivity. We then compare the algorithm with the classical spectral subtraction method and the Qualcomm-ICSI-OGI noise reduction method. We optimize the sound quality in terms of signal-to-noise ratio and provide results on a noisy speech recognition task.
|Title of host publication||Proceedings of the 2007 IEEE Signal Processing Society Workshop, MLSP : Machine Learning for Signal Processing 17|
|Publication status||Published - 2007|
|Event||2007 IEEE International Workshop on Machine Learning for Signal Processing - Thessaloniki, Greece|
Duration: 27 Aug 2007 → 29 Aug 2007
|Workshop||2007 IEEE International Workshop on Machine Learning for Signal Processing|
|Period||27/08/2007 → 29/08/2007|