Name
Abstract | ||
|---|---|---|
Two approaches for modulation spectrum equalization are proposed for robust feature extraction in speech recognition. In both cases the temporal trajectories of the feature parameters are first transformed into the modulation spectrum. In the spectral histogram equalization (SHE) approach, we equalize the histogram of the modulation spectrum for each utterance to a reference histogram obtained from clean training data. In the magnitude ratio equalization (MRE) approach, we equalize the magnitude ratio of lower to higher frequency components on the modulation spectrum to a reference value also obtained from clean training data. Preliminary experimental results performed on the AURORA 2 testing environment indicate that significant performance improvements are achievable with these approaches, when integrated with cepstral mean and variance normalization (CMVN), for all testing sets A, B, and C, all types of noise, for all SNR values. We also show that the approach of magnitude ratio equalization (MRE) offers additional performance improvements when integrated with other more advanced feature normalization approaches such as histogram equalization (HEQ) and higher-order cepstral moment normalization (HOCMN). |
Year | DOI | Venue |
|---|---|---|
2007 | 10.1109/ASRU.2007.4430088 | ASRU |
Keywords | Field | DocType |
modulation spectrum equalization,speech recognition,spectral histogram equalization approach,statistical analysis,snr value,modulation spectrum,aurora 2 testing environment,cepstral mean,magnitude ratio equalization approach,variance normalization,cepstral analysis,feature normalization,feature extraction,temporal filter,robust speech recognition,higher-order cepstral moment normalization,histogram equalization,robust feature extraction,temporal trajectory,higher order,spectrum,indexing terms | Histogram,Normalization (statistics),Equalization (audio),Pattern recognition,Computer science,Cepstrum,Adaptive histogram equalization,Feature extraction,Speech recognition,Cepstral Mean and Variance Normalization,Artificial intelligence,Histogram equalization | Conference |
ISBN | Citations | PageRank |
978-1-4244-1746-9 | 4 | 0.48 |
References | Authors | |
8 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Liang-Che Sun | 1 | 36 | 3.43 |
| Chang-Wen Hsu | 2 | 39 | 2.50 |
| Lin-shan Lee | 3 | 1525 | 182.03 |