Evolutionary process indicators for active IGAs applied to weight tuning in unit selection TTS synthesis

Abstract

Text-to-Speech (TTS) synthesis systems produce speech from an input text. Corpus based or unit selection TTS (US-TTS) are based on retrieving the best set of speech units from a large labelled speech database. To that effect, the unit selection is guided by dynamic programming and a weighted cost function. Several weight tuning approaches have been defined so as to integrate human preferences inthe unit selection process, but with no great success beyond expert-based hand tuning. However, active interactive genetic algorithms (aiGAs) have showed promising results working on the unit selection text-to-speech (US-TTS) weight tuning problem in previous works. aiGAs are an evolution of classic interactive genetic algorithms (IGAs) in terms of reducing fatigue, ambiguity and frustration in user's evaluations. This paper presents a step further in the application of aiGAs to this problem by defining new indicators of the perceptually-based evolutionary process to obtain more reliable weights. The experiments have been conducted on one hour Spanish speech database and using an acoustic plus linguistic cost function.

Efficient and reliable perceptual weight tuning for unit-selection text-to-speech synthesis based on active interactive genetic algorithms: A proof-of-concept

“…This might occur either because a good solution has been found before finishing the tournaments or because the user has become fatigued, being unable to realize the minor variations the last iteration of the evolutionary process typically attain. The stop criterion of the iterative process is determined by considering the evolution of the number of draws with respect to the total number of comparisons of the directed graph (i.e., the certainty ratio) [39]). The last iteration the certainty ratio increases (before decreasing until the end of the iterative process) is considered as the point where the user has converged (see Fig.…”