Sciweavers

Free Online Productivity Tools i2Speak i2Symbol i2OCR iTex2Img iWeb2Print iWeb2Shot i2Type iPdf2Split iPdf2Merge i2Bopomofo i2Arabic i2Style i2Image i2PDF iLatex2Rtf Sci2ools

11

ICASSP
2010
IEEE

favoriteEmaildiscussreport

154views Signal Processing» more ICASSP 2010»

Histogram equalization and noise masking for robust speech recognition

13 years 4 months ago

Histogram equalization and noise masking for robust speech recognition

Download www.esat.kuleuven.be

Mismatch between training and test conditions deteriorates the performance of speech recognizers. This paper investigates the combination of parametric histogram equalization (pHEQ) and noise masking to compensate for the mismatch caused by additive noise. The proposed front-end maps the distribution of the observed power spectrum vectors to a target distribution. The target distribution matches the distribution of the noise free training data except for an artiﬁcially reduced signal-to-noise ratio. Different power spectrum estimation algorithms are used to estimate the noise distribution as used internally by pHEQ more reliably under nonstationary noise conditions. The proposed front-end is evaluated on the Aurora4 database and shows a signiﬁcant improvement w.r.t. mean-normalized Mel-frequency spectral coefﬁcients. Moreover, the performance could be further improved if better estimates of the instantaneous noise power spectrum were available.

Xueru Zhang, Kris Demuynck, Hugo Van Hamme

Real-time Traffic

ICASSP 2010 | Noise Free Training | Power Spectrum | Signal Processing | Target Distribution |

claim paper

Related Content

» The effectiveness of histogram equalization on environmental model adaptation

» Sparse imputation for noise robust speech recognition using soft masks

» Switching Linear Dynamic Models for Noise Robust InCar Speech Recognition

» Mask estimation in nonstationary noise environments for missing feature based robust speec...

» Maximum likelihood adaptation of histogram equalization with constraint for robust speech ...

» Robust speech recognition by integrating speech separation and hypothesis testing

» On noise masking for automatic missing data speech recognition A survey and discussion

» A criterion for the enhancement of timefrequency masks in missing data recognition

» Robust automatic speech recognition with decoder oriented ideal binary mask estimation

Post Info
More Details (n/a)

Added	06 Dec 2010
Updated	06 Dec 2010
Type	Conference
Year	2010
Where	ICASSP
Authors	Xueru Zhang, Kris Demuynck, Hugo Van Hamme

Comments (0)