Enhancements of viterbi search for fast unit selection synthesis

TLDR

Viterbi search in unit‑selection TTS suffers performance degradation when scaling to large corpora required for naturalness. The study aims to accelerate Viterbi search by integrating advanced stopping schemes and pruning thresholds into the baseline algorithm. The authors optimized Viterbi search by adding sophisticated stopping schemes and pruning thresholds, and by using only three intuitive coefficients for configuration, achieving speed‑ups of 6–58× across scenarios while maintaining quality as verified by CCR listening tests. The optimized search enables device‑specific tuning and achieves significant performance gains.

Abstract

The paper describes the optimisation of Viterbi search used in unit selection TTS, since with a large speech corpus necessary to achieve a high level of naturalness, the performance still suffers. To improve the search speed, the combination of sophisticated stopping schemes and pruning thresholds is employed into the baseline search. The optimised search is, moreover, extremely flexible in configuration, requiring only three intuitively comprehensible coefficients to be set. This provides the means for tuning the search depending on device resources, while it allows reaching significant performance increase. To illustrate it, several configuration scenarios, with speed-up ranging from 6 to 58 times, are presented. Their impact on speech quality is verified by CCR listening test, taking into account only the phrases with the highest number of differences when compared to the baseline search.