Graduate School and Research Center in Digital Sciences

Efficient super-wide bandwidth extension using linear prediction based analysis-synthesis

Bachhav, Pramod; Todisco, Massimiliano; Evans, Nicholas

ICASSP 2018, IEEE International Conference on Acoustics, Speech and Signal Processing, April 15-20, 2018, Calgary, Alberta, Canada

Many smart devices now support high-quality speech communication services at super-wide bandwidths. Often, however, speech quality is degraded when they are used with networks or devices which lack super-wideband support. Artificial bandwidth extension can then be used to improve speech quality. While approaches to wideband extension have been reported previously, this paper proposes an approach to super-wide bandwidth extension. The algorithm is based upon a classical source filter model in which spectral envelope and residual error information are extracted from a wideband signal using conventional linear prediction analysis. A form of spectral mirroring is then used to extend the residual error component before an extended super-wideband signal is derived from its combination with the original wideband envelope. Improvements to speech quality are confirmed with both objective and subjective assessments. These show that the quality of super-wideband speech, derived from the bandwidth extension of wideband speech, is comparable to that of speech processed with the standard enhanced voice services (EVS) codec with a bitrate of 13.2kbps. Without the need for statistical estimation of missing super-wideband components, the proposed algorithm is highly efficient and introduces only negligible latency.

Document Doi Bibtex

Title:Efficient super-wide bandwidth extension using linear prediction based analysis-synthesis
Keywords:bandwidth extension, super-wideband, voice quality
Department:Digital Security
Eurecom ref:5502
Copyright: © 2018 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Bibtex: @inproceedings{EURECOM+5502, doi = {}, year = {2018}, title = {{E}fficient super-wide bandwidth extension using linear prediction based analysis-synthesis}, author = {{B}achhav, {P}ramod and {T}odisco, {M}assimiliano and {E}vans, {N}icholas}, booktitle = {{ICASSP} 2018, {IEEE} {I}nternational {C}onference on {A}coustics, {S}peech and {S}ignal {P}rocessing, {A}pril 15-20, 2018, {C}algary, {A}lberta, {C}anada }, address = {{C}algary, {CANADA}}, month = {04}, url = {} }
See also: