Speaker Recognition based on Relative Spectral MethodologyA simple and effective code for RASTA-PLP Speaker Identification. | |
Download |
Speaker Recognition based on Relative Spectral Methodology Ranking & Summary
Advertisement
- License:
- Freeware
- Publisher Name:
- Luigi Rosa
- Operating Systems:
- Windows All
- File Size:
- 39 KB
Speaker Recognition based on Relative Spectral Methodology Tags
Speaker Recognition based on Relative Spectral Methodology Description
The Perceptual Linear Predicative (PLP) speech analysis technique is based on the short-term spectrum of speech. Even though the short-term spectrum of speech is subsequently modified by several psychophysically based spectral transformations, the PLP technique (just like most other short-term spectrum based techniques), is vulnerable when the short-term spectral values are modified by the frequency response of the communication channel. Human speech perception seems to be less sensitive to such steady-state spectral factors. We have implemented the RelAtive SpecTrAl (RASTA) methodology which makes PLP (and possibly also some other short-term spectrum based techniques) more robust to linear spectral distortions. Experimental results using telephone-quality isolated digits and hìgh-quality continuous speech show significant improvements in error rate. Give it a try to see what it's really capable of!
Speaker Recognition based on Relative Spectral Methodology Related Software