The clear speech intelligibility benefit for text-to-speech voices: Effects of speaking style and visual guise

Aoki, Nicholas B.; Cohn, Michelle; Zellou, Georgia

doi:10.1121/10.0010274

Cited by 17 publications

(20 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This availability of tools to leverage AI speech could facilitate new advances in research and clinical disciplines, including hearing science, audiology, and speech pathology, in which creating and norming human-spoken speech materials for diagnostic and rehabilitative purposes is an ongoing challenge, and could potentially be substituted by AI speech synthesis. However, research thus far has focused on technology development towards human-like voices and the degree to which AI speech is perceived by young, normal-hearing listeners (e.g., Govender et al, 2019b;Govender et al, 2019a;Aoki et al, 2022). How older adults perceive and experience AI speech is unclear and topic of the current study.…”

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 95%

“…A few recent studies have compared perception of human speech to AI-based synthesized speech in a variety of contexts Aoki et al, 2022). For example, studies have investigated the experienced valence of human vs AI speech or how humans produce speech when talking to voice-AI systems compared to humans Zellou et al, 2021;Cohn et al, 2022).…”

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 99%

“…For example, studies have investigated the experienced valence of human vs AI speech or how humans produce speech when talking to voice-AI systems compared to humans Zellou et al, 2021;Cohn et al, 2022). Most relevant for the current work, a few studies have investigated whether speech intelligibility differs between synthesized speech and human-spoken speech (Govender et al, 2019b;Govender et al, 2019a;Aoki et al, 2022). Intelligibility for speech masked by background sound has been suggested to be lower for synthesized speech compared to human-spoken speech (Cooke et al, 2013;Simantiraki et al, 2018;Aoki et al, 2022).…”

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 99%

“…Most relevant for the current work, a few studies have investigated whether speech intelligibility differs between synthesized speech and human-spoken speech (Govender et al, 2019b;Govender et al, 2019a;Aoki et al, 2022). Intelligibility for speech masked by background sound has been suggested to be lower for synthesized speech compared to human-spoken speech (Cooke et al, 2013;Simantiraki et al, 2018;Aoki et al, 2022). However, the most recent AI speech synthesizers have not been available at the time some of the works were conducted, thus limiting this conclusion somewhat (Cooke et al, 2013;Simantiraki et al, 2018).…”

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 99%

“…However, the most recent AI speech synthesizers have not been available at the time some of the works were conducted, thus limiting this conclusion somewhat (Cooke et al, 2013;Simantiraki et al, 2018). Moreover, only a small number of masking levels (N=1-3) have often been used in previous work (e.g., Cooke et al, 2013;Govender et al, 2019b;Govender et al, 2019a;Aoki et al, 2022). Presentation of a low number of masking levels can limit interpretations about whether AI speech is less intelligible than human speech, because intelligibility differences can be related to differences in speech-perception thresholds or speech-perception variability (i.e., the steepness of the relation between masking levels and intelligibility).…”

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 99%

See 4 more Smart Citations

The perception of artificial-intelligence (AI) based synthesized speech in younger and older adults

Herrmann

2022

Preprint

View full text Add to dashboard Cite

Artificial intelligence (AI) based synthesized speech has become almost human-like, ubiquitous in everyday live (e.g., smart phones, grocery self-checkouts), and relatively easy to synthesize. This opens opportunities to use AI speech in research and clinical areas, such as hearing sciences, audiology, and speech pathology, where recordings of speech materials by voice actors can be time- and cost-intensive. However, much research thus far has focused on technological developments towards more human-like voices evaluated by younger adults. How older adults perceive AI speech is unclear. Using Google’s Wavenet text-to-speech synthesizer, the current study explores whether AI speech can be used to investigate common speech-in-noise perception phenomena in younger and older adults. Speech intelligibility was recorded for human speech and synthesized speech masked by a modulated or an unmodulated multi-talker babble noise. For both human and AI speech, speech intelligibility was better for the modulated than the unmodulated masker (masking release), and this masking-release benefit was reduced in older adults. Release from masking effects were comparable between human and AI speech, suggesting that modern AI speech could be useful for hearing and speech research. The data further suggest that older adults recognize the presentation of AI speech less frequently, rate AI speech as more natural, and are less able to discriminate between human and AI speech compared to younger adults. Research on speech perception in older adults may thus especially benefit from modern AI-based synthesized speech because, to them, AI speech feels much like spoken by a human.

show abstract

Section: Introduction Introduction Introduction Introductionmentioning

confidence: 95%