Name
Title | ||
|---|---|---|
Advanced parallel combined Gaussian mixture model based feature compensation integrated with iterative channel estimation |
Abstract | ||
|---|---|---|
•Effective feature compensation for speech recognition in noise and channel distortion.•Employs Parallel Combined Gaussian Mixture Model (PCGMM).•Evaluation uses objective measures including STNR, PESQ, and speech recognition.•Show +9.77% and +15.77% relative avg. WER improvement vs. ETSI AFE standard. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1016/j.specom.2015.07.008 | Speech Communication |
Keywords | Field | DocType |
Robust speech recognition,Feature compensation,Channel estimation,Model combination,PCGMM | Speech corpus,Background noise,Pattern recognition,Computer science,Voice activity detection,A priori and a posteriori,Communication channel,Speech recognition,Artificial intelligence,Distortion,Mixture model,PESQ | Journal |
Volume | Issue | ISSN |
73 | C | 0167-6393 |
Citations | PageRank | References |
1 | 0.35 | 27 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Wooil Kim | 1 | 120 | 16.95 |
| John H. L. Hansen | 2 | 3215 | 365.75 |