Sparsification of voice data using Discrete Rajan Transform and its applications in speaker recognition

This paper proposes a novel technique of sparsing speech data and compressing it in spectral domain. Discrete Rajan Transform is applied to voice data and the spectrum is sparsed by retaining the first component CPI (Cumulative Point Index) of the spectrum and forcing the other spectral components to zero. Thus the spectrum could be compressed to a maximum of 12.5% of the original data. As and when required the compressed spectrum could be synthesized using Inverse Discrete Rajan Transform and the reconstructed speech data analyzed for speaker recognition. Speaker recognition accuracy to a maximum of 93.5% has been obtained in this case.