A Low Bit Rate Scalable CWI Coder based on Wavelet Transform

This paper presents a scalable characteristic waveform interpolation (CWI) speech coder based on wavelet transform (WT) at low bit rate ranging from 1.8 kbit/s to 3.6 kbit/s. The characteristic waveform surface is decomposed into a series of reduced time resolution surfaces using B-spline biorthogonal linear phase filter banks. The common coding parameters form the basic layer at 1.8 kbit/s that can produce an acceptable speech quality. The different wavelet decomposition surfaces form different enhancement layers and improve the speech quality successively with efficient quantization. Subjective listening test showed that this scalable WT-CWI coder performs well at low bit rate.

[1]  Joe F. Chicharo,et al.  A new waveform interpolation coding scheme based on pitch synchronous wavelet transform decomposition , 2000, IEEE Trans. Speech Audio Process..

[2]  Jerry D. Gibson,et al.  Structures for SNR scalable speech coding , 2006, IEEE Transactions on Audio, Speech, and Language Processing.

[3]  Allen Gersho,et al.  Variable-dimension vector quantization , 1996, IEEE Signal Process. Lett..

[4]  Jerry D. Gibson,et al.  SNR and bandwidth scalable speech coding , 2002, 2002 IEEE International Symposium on Circuits and Systems. Proceedings (Cat. No.02CH37353).

[5]  Eddie L. T. Choy,et al.  Waveform Interpolation Speech Coder at 4 kb/s , 1998 .

[6]  Balázs Kövesi,et al.  A scalable speech and audio coding scheme with continuous bitrate flexibility , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.