Title
Dynamic bit allocation in CELP excitation coding
Abstract
The excitation coding, i.e. the LTP (Lon Term Predictor) and the innovation coding, requires afarge part of the overall bit rate in a CELP coder. In this paper we ropose a method to reduce the excitation coding bit rate. The fact that the pitch varies only slowly during voiced segments of speech can be exploited to design powerful dynamic bit allocation schemes for the excitation sequence. The bit allocation is determined by two methods. In method one, the LTP index is Huffman coded. This makes the LTP code book require only a small number of bits during speech segments with stable pitch frequency, i.e. voiced segments. In method two, a high rate approximation to assign various numbers of innovation code words for each LTP index is derived. To avoid variable transmission rate, the total bit rate is held constant for each sub frame. As a complement to dynamic bit allocation, a search method for the LTP index is developed, that takes into account the number of innovation code words assigned to each LTP index, in the search for an optimal LTP sequence. Simulations are included that shows that with these methods the bit rate can be decreased by 400 bits/s with no changes in speech quality.
Year
DOI
Venue
1993
10.1109/ICASSP.1993.319261
ICASSP
Keywords
Field
DocType
ltp index,powerful dynamic bit allocation,bit allocation,innovation code word,bit rate,optimal ltp sequence,ltp code book,overall bit rate,celp excitation coding,excitation coding bit rate,dynamic bit allocation,speech coding,code excited linear prediction,huffman codes,huffman coding,linear predictive coding,frequency,indexation,information theory,speech segmentation,data compression
Speech coding,Computer science,Bit field,Huffman coding,Artificial intelligence,Code word,Linear predictive coding,Code-excited linear prediction,Pattern recognition,Algorithm,Speech recognition,Harmonic Vector Excitation Coding,Data compression
Conference
Volume
ISSN
ISBN
2
1520-6149
0-7803-0946-4
Citations 
PageRank 
References 
2
0.57
2
Authors
2
Name
Order
Citations
PageRank
Thomas Eriksson130.95
Johan Sjöberg231.61