Analysis-by-Synthesis Speech Coding2008-01-010 PRE AI DOI:10.1007/978-3-540-49127-9_17OriginalOriginal request for helpShareSaveAbstract En