Title | ||
---|---|---|
Robust parameters for speech recognition based on subband spectral centroid histograms |
Abstract | ||
---|---|---|
In this paper we propose a new speech parameterization frame- work that efficiently combines frequency and magnitude infor- mation from the short-term power spectrum of speech. This is achieved through computation of subband spectral centroid his- tograms (SSCH). Relationship between the proposed method and auditory based speech parameterization methods is dis- cussed. An experimental study on an automatic speech recogni- tion task has shown that the proposed method outperforms the conventional speech front-ends in presence of different types of additive noise, while it performs comparably in the noise-free conditions. In the case of car noise, our method also outper- forms the computationally expensive auditory based methods, while having simplicity and low computational cost similar to the conventional front-ends. |
Year | Venue | Keywords |
---|---|---|
2001 | INTERSPEECH | speech recognition,front end,power spectrum |
Field | DocType | Citations |
Histogram,Magnitude (mathematics),Spectral centroid,Pattern recognition,Parametrization,Computer science,Speech recognition,Spectral density,Artificial intelligence,Car noise,Computation | Conference | 1 |
PageRank | References | Authors |
0.44 | 7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Bojana Gajic | 1 | 15 | 2.53 |
Kuldip K. Paliwal | 2 | 1890 | 154.90 |