Name
Abstract | ||
|---|---|---|
Recently we introduced a new and simple denoising algorithm, called SPACE, that yielded promising preliminary results in noise robust speech recognition. SPACE is essentially based on GMM modeling of clean an noisy speech. In this paper, we evaluate the performance of SPACE on Aurora2 and show that they are globally not satisfactory, essentially because the Gaussian correspondence assumption is not verified. We then propose a new training procedure for the GMMs that achieves a better Gaussian correspondence. We further develop a simple adaptation algorithm to handle unknown environments that preserves the Gaussian correspondence. We evaluate the new denoising algorithm on Aurora2. The results show that it outperforms the multistyle models, sometimes significantly, on the three test sets of Aurora2 |
Year | DOI | Venue |
|---|---|---|
2006 | 10.1109/ICASSP.2006.1660072 | ICASSP (1) |
Keywords | Field | DocType |
signal denoising,space denoising algorithm,speech recognition,noise robust speech recognition,gaussian processes,gmm modeling,stereo-based piecewise affine compensation for environments,aurora2,gaussian correspondence assumption,gaussian mixture model,automatic speech recognition,acoustic noise,noise reduction,gaussian noise,hidden markov models,testing | Denoising algorithm,Pattern recognition,Computer science,Gaussian,Artificial intelligence,Gaussian process,Gaussian noise | Conference |
Volume | ISSN | ISBN |
1 | 1520-6149 | 1-4244-0469-X |
Citations | PageRank | References |
2 | 0.41 | 2 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Christophe Cerisara | 1 | 115 | 19.74 |
| Khalid Daoudi | 2 | 145 | 23.68 |