An evaluation of text-to-speech synthesizers in the foreign language classroom: learners’ perceptions

Abstract

As stated in Cardoso, Smith, and Garcia Fuentes (2015), second language researchers and practitioners have explored the pedagogical capabilities of Text-To-Speech synthesizers (TTS) for their potential to enhance the acquisition of writing (e.g. Kirstein, 2006), vocabulary and reading (e.g. Proctor, Dalton, & Grisham, 2007), and pronunciation (e.g. Cardoso, Collins, & White, 2012). Despite their demonstrated effectiveness, there is a need for up-to-date formal evaluations of TTS systems, specifically for their potential to promote the ideal conditions under which languages are acquired, particularly in an English as a Foreign Language (EFL) environment, as suggested by Cardoso, Smith, and Garcia Fuentes (2015). This study evaluated a modern English TTS system in an EFL context in Brazil, at a number of levels, including speech quality, opportunity to focus on form, and learners' cognitive processing of TTS-generated texts. Fifteen Brazilian EFL learners participated in the study in which they listened to both human and TTSproduced speech samples while performing the abovementioned tasks. Semistructured interviews were used to collect data about participants' perceptions of the technology. We report an analysis of these interviews, which indicate that EFL learners have overall positive attitudes towards the pedagogical use of TTS, and that they would like to use the technology as a learning tool.

Using Challenges to Enhance a Learning Game for Pronunciation Training of English as a Second Language

“…Research on the quality of TTS has led Google to assert that deep neural network (DNN) technology already produces near-human speech [56]. As mentioned earlier in the introduction, despite a certain amount of controversy concerning the pedagogical use of TTS systems in L2 teaching, there is empirical evidence that supports its applicability [57]- [59].…”

Estoñol, a computer-assisted pronunciation training tool for Spanish L1 speakers to improve the pronunciation and perception of Estonian vowels

“…In particular, ASR-based CAPT systems present several advantages, such as dynamic evaluation, more intensive practice, immediate-automated feedback, anxiety-free context, individualized feedback, and opportunities for repair (Neri et al 2002(Neri et al , 2006. Even though old TTS systems generated a great deal of controversy when used as a pedagogical tool, recent research in speech synthesis has reported some benefits in terms of naturalness, comprehensibility, accuracy, and intelligibility (Smith et al 2015 andBione et al 2016).…”

Proceedings of the Workshop on NLG for Human–Robot Interaction

“…In the context of language acquisition, the quality of the synthesised speech may be of importance, as (young) learners have been shown to attend to non-verbal cues present in the speech signal when presented with a novel language (e.g., Dominey and Dodane, 2004). Although the effects of speech synthesis quality on learners' perceptions have previously been studied for computerassisted language learning (e.g., Bione et al, 2016;Handley, 2009;Kang et al, 2008), whether poor quality speech synthesis impedes the efficacy of language acquisition has not been unequivocally established.…”