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

Abstract

Unit-selection speech synthesis is one of the current corpus-based text-tospeech synthesis techniques. The quality of the generated speech depends on the accuracy of the unit selection process, which in turn relies on the cost function definition. This function should map the user perceptual preferences when selecting synthesis units, which is still an open research issue. This paper proposes a complete methodology for the tuning of the cost function weights by fusing the human judgments with the cost function, through efficient and reliable interactive weight tuning. To that effect, active interactive genetic algorithms (aiGAs) are used to guide the subjective weight adjustments. The application of aiGAs to this process allows mitigating user fatigue and frustration by improving user consistency. However, it is still unfeasible to subjectively adjust the weights of the whole corpus units (diphones and triphones in this work). This makes it mandatory to perform unit clustering before conducting the tuning process. The aiGA-based weight tuning proposal is evaluated in a small speech corpus as a proof-of-concept and results in more natural synthetic speech when compared to previous objective

A unit selection text-to-speech-and-singing synthesis framework from neutral speech: proof of concept

“…This block retrieves the units that minimise the prosodic, linguistic, and concatenation costs (see [42] for more details). The weights for the prosodic target and concatenation subcosts were perceptually tuned by means of active interactive genetic algorithms for speech synthesis purposes [44].…”

Low Footprint High Intelligibility Malay Speech Synthesizer Based on Statistical Data

“…Therefore, this method involves less signal processing or no signal processing. Unit selection method is popular attributable to its high intelligibility and naturalness of output speech (Alias et al, 2011). However, demands larger database for better quality (BarraChicote et al, 2010).…”