Kazutoshi Yasunaga

A candidate for the ITU-T 4 kbit/s speech coding standard

This paper presents the 4 kbit/s speech coding candidate submitted by AT&T, Conexant, Deutsche Telekom, France Telecom, Matsushita, and NTT for the ITU-T 4 kbit/s selection phase. The algorithm was developed jointly based on the qualification version of Conexant. This paper focuses on the development carried out during the collaboration in order to narrow the gap to the requirements in an attempt to provide toll quality at 4 kbit/s. This objective is currently being verified in independent subjective tests coordinated by ITU-T and carried out in multiple languages. Subjective tests carried out during the development indicate that the collaboration work has been successful in improving the quality, and that meeting a majority of the requirements in the extensive selection phase test is a realistic goal.

Dispersed-pulse codebook and its application to a 4 kb/s speech coder

This paper presents a dispersed-pulse codebook for CELP coder. This codebook generates an excitation vector by convoluting dispersion vectors with signed pulses in an algebraic codevector. The dispersion vectors are obtained through training so the coding distortion to be reduced. An objective evaluation result shows that the coding distortion with this codebook is smaller than that with an algebraic codebook. The dispersed-pulse codebook is applied to a 4 kb/s CELP coder. Subjective evaluation results show that: (1) the fundamental performance of the 4 kb/s coder is equivalent to that of G.726 32 kb/s coder, and (2) the performances of the 4k b/s coder under some error and background noise conditions are equivalent to those of G.729 8 kb/s coder.

Noise post processing based on a stationary noise generator

The paper presents a newly developed noise post-processing (NPP) algorithm for improving subjective performance of low bit-rate code excited linear prediction (CELP) decoding. The algorithm is based on a stationary noise generator and improves subjective quality of a noisy signal input. A backward adaptive detector defines noisy input signal segments from decoded line spectral frequency (LSF), energy, and pitch parameters. By using LSF and energy parameters from the past, the noise generator estimates and produces stationary noise. The algorithm is independent of the main codec algorithm and is applicable to any CELP decoder. When applied to G.729 and G.723.1, our test results show that the algorithm improves subjective performance at around 0.2-0.8 in degradation category rating (DCR) mean opinion score (DMOS) under several background noise conditions.

A high quality 4-kbit/s speech coding algorithm based on MDP-CELP

This paper presents a high quality 4-kbit/s speech coding algorithm based on the CELP algorithm. This coder operates on speech frames of 20 ms. Its main features are as follows: modification of a target vector for LSP-VQ which improves the subjective quality of the coded speech signal, a dispersed-pulse structure which improves pulse-based fixed-codebook performance at a low bit rate, a high performance fixed-codebook consists of a partial algebraic codebook and a random codebook, backward mode switching which improves fixed codebook performance for unvoiced and stationary noise regions, a stationary noise generator which significantly improves the subjective quality under background noise conditions. Subjective testing indicates that the quality of this coding algorithm is better than that of ITU-T G.726 (32-kbit/s ADPCM) under error-free and noise-free conditions and equivalent to that of ITU-T G.729 (8-kbit/s CS-ACELP) under some background noise conditions.