Predictive multiple-scale lattice VQ for LSF quantization

Abstract

This paper introduces a new lattice quantization scheme, the multiple-scale lattice vector quantization (MSLVQ), based on the truncation of the D + 10 lattice. The codebook is composed of several copies of the truncated lattice scaled with different scaling factors. A fast nearest neighbor search is introduced. We compare the performance of predictive MSLVQ for quantization of LSF coefficients with the quantization technique used in the codec G.729 and show the better performance of our method in terms of spectral distortion. The MSLVQ scheme achieves the transparent quality at 21 bits/frame.

New Directions in Lattice Based Lossy Compression

“…In (Solé, 1993) the theta functions have been generalized for the norm ℓ p and in (Moureaux et al, 1995), (Barlaud et al, 1993) for weighted ℓ 2 norms. In (Vasilache et al, 1999) the theta series approach is used to count the lattice points on spherical (ℓ 2 norm) shells and generalized to pyramidal (ℓ 1 norm) shells.…”