A Novel Sinusoidal Speech Codec Using Multiple Descriptions

Robust and flexible speech codecs are more and more required by speech communication over unreliable channels such as Internet. In this paper, a novel multiple description (MD) sinusoidal speech codec is proposed. This codec is based on sinusoidal and equivalent rectangular bands (ERB) noise model. It can provide relatively high transmission reliability as well as good coding efficiency. And the lost packet doesn't affect the state recovery of this state-less MD codec. Therefore it is very suitable to unreliable and band limited channel such as Internet

[1]  Michael M. Goodwin.  Residual modeling in music analysis-synthesis , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[2]  Ajay Ingle,et al.  DPCM system design for diversity systems with applications to packetized speech , 1995, IEEE Trans. Speech Audio Process..

[3]  Jerry D. Gibson,et al.  A multiple description speech coder based on AMR-WB for mobile ad hoc networks , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[4]  Mark J. T. Smith,et al.  Speech analysis/synthesis and modification using an analysis-by-synthesis/overlap-add sinusoidal model , 1997, IEEE Trans. Speech Audio Process..

[5]  Nuggehally Sampath Jayant,et al.  Effects of Packet Losses in Waveform Coded Speech and Improvements Due to an Odd-Even Sample-Interpolation Procedure , 1981, IEEE Trans. Commun..

[6]  Benjamin W. Wah,et al.  LSP-based multiple-description coding for real-time low bit-rate voice over IP , 2005, IEEE Transactions on Multimedia.