On the impact of dysarthric speech on contemporary ASR cloud platforms

Russis, Luigi De; Corno, Fulvio

doi:10.1007/s40860-019-00085-y

Cited by 40 publications

(22 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the obvious barriers that some users with disabilities can encounter by interacting with voice assistants is related to speech impairments [ 13 ] that are a frequent secondary consequence of motor disorders [ 14 ]. Although most voice assistants exploit machine learning algorithms to adapt to the user and increase their speech recognition accuracy over time [ 15 , 16 ], these systems are still designed and developed for people with clear and intelligible speech.…”

Section: Introductionmentioning

confidence: 99%

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study

et al. 2020

View full text Add to dashboard Cite

Background Voice assistants allow users to control appliances and functions of a smart home by simply uttering a few words. Such systems hold the potential to significantly help users with motor and cognitive disabilities who currently depend on their caregiver even for basic needs (eg, opening a door). The research on voice assistants is mainly dedicated to able-bodied users, and studies evaluating the accessibility of such systems are still sparse and fail to account for the participants’ actual motor, linguistic, and cognitive abilities. Objective The aim of this work is to investigate whether cognitive and/or linguistic functions could predict user performance in operating an off-the-shelf voice assistant (Google Home). Methods A group of users with disabilities (n=16) was invited to a living laboratory and asked to interact with the system. Besides collecting data on their performance and experience with the system, their cognitive and linguistic skills were assessed using standardized inventories. The identification of predictors (cognitive and/or linguistic) capable of accounting for an efficient interaction with the voice assistant was investigated by performing multiple linear regression models. The best model was identified by adopting a selection strategy based on the Akaike information criterion (AIC). Results For users with disabilities, the effectiveness of interacting with a voice assistant is predicted by the Mini-Mental State Examination (MMSE) and the Robertson Dysarthria Profile (specifically, the ability to repeat sentences), as the best model shows (AIC=130.11). Conclusions Users with motor, linguistic, and cognitive impairments can effectively interact with voice assistants, given specific levels of residual cognitive and linguistic skills. More specifically, our paper advances practical indicators to predict the level of accessibility of speech-based interactive systems. Finally, accessibility design guidelines are introduced based on the performance results observed in users with disabilities.

show abstract

Section: Introductionmentioning

confidence: 99%

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Notably, this apparent data bias is not limited to commercial ASR systems, as Mozilla's open-source DeepSpeech system trained on their crowdsourced CommonVoice corpus also performs significantly worse for AAE and Indian English than Mainstream US English (Martin and Tang, 2020;Meyer et al, 2020). Other work has focused on the use of ASR as an assistive technology and found that most major systems perform poorly for Deaf and hard of hearing (Glasser, 2019), and dysarthric users (De Russis and Corno, 2019;Young and Mihailidis, 2010).…”

Section: Predictive Bias In Asrmentioning

confidence: 99%

Language technology practitioners as language managers: arbitrating data bias and predictive bias in ASR

Markl¹,

McNulty²

2022

Preprint

View full text Add to dashboard Cite

Despite the fact that variation is a fundamental characteristic of natural language, automatic speech recognition systems perform systematically worse on non-standardised and marginalised language varieties. In this paper we use the lens of language policy to analyse how current practices in training and testing ASR systems in industry lead to the data bias giving rise to these systematic error differences. We believe that this is a useful perspective for speech and language technology practitioners to understand the origins and harms of algorithmic bias, and how they can mitigate it. We also propose a re-framing of language resources as (public) infrastructure which should not solely be designed for markets, but for, and with meaningful cooperation of, speech communities.

show abstract

“…One of the obvious barriers that some users with disabilities can encounter by interacting with voice assistants is related to speech impairments [13] that are a frequent secondary consequence of motor disorders [14]. Although most voice assistants exploit machine learning algorithms to adapt to the user and increase their speech recognition accuracy over time [15,16], these systems are still designed and developed for people with clear and intelligible speech.…”

Section: Introductionmentioning

confidence: 99%

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study (Preprint)

Masina¹,

Orso²,

Pluchino³

et al. 2020

Preprint

View full text Add to dashboard Cite

BACKGROUND Voice assistants allow users to control appliances and functions of a domotic house by simply uttering a few words. Such systems hold the potential to significantly help users with motor and cognitive disabilities who currently depend on their caregiver even for basic needs, e.g., opening a door. The research on voice assistants is mainly dedicated to abled-bodied users, and studies evaluating the accessibility of such systems are still sparse and fail to account for the participants’ actual cognitive and linguistic abilities. OBJECTIVE In the present work, we aimed to investigate whether cognitive and/or linguistic functions could predict the user’s performance in operating an off-the-shelf voice assistant, namely Google Home. METHODS A group of users with disabilities (N=16) was involved in a living laboratory and was asked to interact with the system. Besides collecting data regarding their performance and their experience with the system, we assessed their cognitive and linguistic skills using standardized inventories. The identification of predictors (cognitive and/or linguistic) capable to account for an efficient interaction with the voice assistant was investigated by performing multiple linear regression models. The best model was identified by adopting a selection strategy based on the Akaike Information Criterion (AIC). RESULTS The best model (AIC = 127.81) shows that, among several cognitive and linguistic predictors, the Mini-Mental State Examination (MMSE) and the ability to repeat sentences (Robertson Dysarthria Profile) were crucial to predict the user’s performance. CONCLUSIONS Our findings show that cognitive and linguistic skills are involved and co-participate in the effective interaction with voice assistants. Furthermore, the present study provides design implications based on the mistakes observed during the interaction with voice assistants and advances practical indicators to predict the level of accessibility of voice assistants.

show abstract

On the impact of dysarthric speech on contemporary ASR cloud platforms

Cited by 40 publications

References 18 publications

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study

Language technology practitioners as language managers: arbitrating data bias and predictive bias in ASR

Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study (Preprint)

Contact Info

Product

Resources

About