Backward adaptation for low delay vector excitation coding of speech at 16 kbit/s

To attain a very-low-delay speech coder at 16 kb/s while maintaining a quality acceptable for the public switched telephone network, low delay vector excitation coding (LD-VXC) is introduced. Backward adaptation is used to track the spectral characteristics of the signal without requiring any buffering of the input speech, thereby allowing a very low delay to be achieved in an analysis-by-synthesis structure. The algorithm differs markedly from conventional VXC or CELP (code-excited linear prediction) coders due to the use of backward adaptive linear prediction for modeling the time-varying short- and long-term correlation of speech. The LD-VXC coder provides very good speech quality at 16 kb/s, moderate complexity, a delay of under 2 ms, and a gentle degradation of quality with transmission errors. The algorithm was submitted to the CCITT as a candidate for a future 16-kb/s speech coding standard.<<ETX>>

[1]  Allen Gersho,et al.  Gain-Adaptive Vector Quantization with Application to Speech Coding , 1987, IEEE Trans. Commun..

[2]  K. Zeger,et al.  Zero redundancy channel coding in vector quantisation , 1987 .

[3]  Manfred R. Schroeder,et al.  Code-excited linear prediction(CELP): High-quality speech at very low bit rates , 1985, ICASSP '85. IEEE International Conference on Acoustics, Speech, and Signal Processing.

[4]  Allen Gersho,et al.  Real-time vector APC speech coding at 4800 bps with adaptive postfiltering , 1987, ICASSP '87. IEEE International Conference on Acoustics, Speech, and Signal Processing.

[5]  V. Cuperman,et al.  Backward pitch prediction for low-delay speech coding , 1989, IEEE Global Telecommunications Conference, 1989, and Exhibition. 'Communications Technology for the 1990s and Beyond.

[6]  V. Cuperman,et al.  A vector ADPCM analysis-by-synthesis configuration for 16 kbit/s speech coding , 1988, IEEE Global Telecommunications Conference and Exhibition. Communications for the Information Age.

[7]  Allen Gersho,et al.  Real-time vector excitation coding of speech at 4800 bps , 1987, ICASSP '87. IEEE International Conference on Acoustics, Speech, and Signal Processing.