Name
Title | ||
|---|---|---|
Regularized minimum variance distortionless response-based cepstral features for robust continuous speech recognition |
Abstract | ||
|---|---|---|
•We study the low-variance and robust features for speech recognition system on the AURORA-4 corpus.•We propose to compute cepstral features from a regularized MVDR (RMVDR) spectral estimates, denoted as RMVDR-based Cepstral Coefficient (RMCC) features.•A sigmoid-shape auditory domain weighting rule is proposed for speech spectrum enhancement and incorporated in to the RMCC framework.•We incorporate the medium duration power bias subtraction (MDPBS) method in to the RMCC framework.•Two robust front-ends are proposed, robust RMCC (RRMCC) and Normalized RMCC (NRMCC) for speech recognition. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1016/j.specom.2015.07.007 | Speech Communication |
Keywords | Field | DocType |
Speech recognition,Robust feature extraction,Regularized MVDR,ASE,Feature normalization,Multi-condition training | Mel-frequency cepstrum,Minimum-variance unbiased estimator,Pattern recognition,Computer science,Cepstrum,Word error rate,Feature extraction,Robustness (computer science),Speech recognition,Spectral density,Artificial intelligence,Estimator | Journal |
Volume | Issue | ISSN |
73 | C | 0167-6393 |
Citations | PageRank | References |
2 | 0.44 | 30 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| jahangir alam | 1 | 320 | 38.69 |
| Patrick Kenny | 2 | 2700 | 214.80 |
| Douglas D. O'Shaughnessy | 3 | 398 | 84.79 |