Proceedings ArticleDOI
Single stage spectral quantization at 20 bits
Per Hedelin
- pp 525-528
Reads0
Chats0
TLDR
The authors demonstrate that high quality encoding of speech spectra is feasible around 20 bits/spectrum with a single-stage approach and suggests a straightforward tree-based procedure that can be employed for the codebook search such that the computational complexity is acceptable.Abstract:
Efficient coding of spectral parameters constitutes a major concern for speech compression. It is generally agreed that a single-stage vector quantizer, operating in the log-spectral domain would give superior performance, as compared to any other block coding procedure. Such a coder has, however, been considered impractical due to its complexity. The authors demonstrate that high quality encoding of speech spectra is feasible around 20 bits/spectrum with a single-stage approach. The experimental results are accurately predicted by a theoretical performance analysis. The author suggests a straightforward tree-based procedure that can be employed for the codebook search such that the computational complexity is acceptable. >read more
Citations
More filters
Book
Digital speech processing, synthesis, and recognition
TL;DR: This paper presents principal characteristics of speech speech production models speech analysis and analysis-synthesis systems linear predictive coding (LPC) analysis speech coding speech synthesis speech recognition future directions of speech processing.
Journal ArticleDOI
Vector quantization based on Gaussian mixture models
Per Hedelin,Jan Skoglund +1 more
TL;DR: It is found that an optimal single-stage VQ can operate at approximately 3 bits less than a state-of-the-art LSF-based 2-split VQ.
Journal ArticleDOI
Interframe LSF quantization for noisy channels
TL;DR: By combining an interframe quantizer and a memoryless "safety-net" quantizer, it is demonstrated that the advantages of both quantization strategies can be utilized, and the performance for both noiseless and noisy channels improves.
Journal ArticleDOI
Optimization of lattices for quantization
Erik Agrell,Thomas Eriksson +1 more
TL;DR: The new nine- and ten-dimensional lattices suggest that Conway and Sloane's (1993) conjecture on the duality between the optimal lattices for packing and quantization might be false.
Journal ArticleDOI
Recursive coding of spectrum parameters
J. Samuelsson,Per Hedelin +1 more
TL;DR: It is shown theoretically that 16 bits are needed to achieve an average SD of 1 dB when quantizing ten-dimensional (10-D) spectrum vectors using a first-order recursive scheme and validated in experiments, and how to approximate the SD with an L/sub 2/-norm measure.
References
More filters
Journal ArticleDOI
Efficient vector quantization of LPC parameters at 24 bits/frame
Kuldip K. Paliwal,B. Atal +1 more
TL;DR: It is shown that the split vector quantizer can quantize LPC information in 24 bits/frame with an average spectral distortion of 1 dB and less than 2% of the frames having spectral distortion greater than 2 dB.
Book
Digital speech processing, synthesis, and recognition
TL;DR: This paper presents principal characteristics of speech speech production models speech analysis and analysis-synthesis systems linear predictive coding (LPC) analysis speech coding speech synthesis speech recognition future directions of speech processing.
Proceedings ArticleDOI
Robust vector quantization in spectral coding
R. Hagen,Per Hedelin +1 more
TL;DR: Using an LSP (line spectrum pair)-representation for the spectrum, it is demonstrated that short block codes are feasible, thus allowing for compact storage of the code-book and secure channel robustness while allowing for efficient design, storage, and handling of the vector quantizer.
Proceedings ArticleDOI
A robust single-stage vq for spectral coding
R. Hagen,Per Hedelin +1 more
TL;DR: A method for design of a single-stage spectral vector quantizer combining the features: efficient storage, fast search, reasonable training complexity, and robustness against channel errors is presented.