Name
Abstract | ||
|---|---|---|
Many techniques for complex speech processing such as denoising and deconvolution, time/frequency warping, multiple speaker separation, and multiple microphone analysis operate on sequences of short-time power spectra (spectrograrns), a representation which is often well-suited to these tasks. However, a significant problem with algorithms that manipulate spectrograms is that the output spectrogram does not include a phase component, which is needed to create a time-domain signal that has good perceptual quality. Here we describe a generative model of time-domain speech signals and their spectrograms, and show how an efficient optimizer can be used to find the maximum a posteriori speech signal, given the spectrogram. In contrast to techniques that alternate between estimating the phase and a spectrally-consistent signal, our technique directly infers the speech signal, thus jointly optimizing the phase and a spectrally-consistent signal. We compare our technique with a standard method using signal-to-noise ratios, but we also provide audio files on the web for the purpose of demonstrating the improvement in perceptual quality that our technique offers. |
Year | Venue | Keywords |
|---|---|---|
2003 | ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 16 | speech processing,time domain,time frequency,signal to noise ratio |
Field | DocType | Volume |
Noise reduction,Speech processing,Computer science,Deconvolution,Artificial intelligence,Image warping,Pattern recognition,Spectrogram,Speech recognition,Maximum a posteriori estimation,Machine learning,Microphone,Generative model | Conference | 16 |
ISSN | Citations | PageRank |
1049-5258 | 9 | 1.00 |
References | Authors | |
5 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kannan Achan | 1 | 425 | 35.52 |
| Sam T. Roweis | 2 | 4556 | 497.42 |
| Brendan J. Frey | 3 | 3637 | 404.51 |