Since the University’s founding, researchers have worked to create new technologies, practices and ideas. Every time someone uses a modern internet browser, turns on an LED light bulb or sends an email, they are witnessing the results of University research. 

Researchers at Illinois have also helped to innovate upon the ways that people live with disabilities since Disability Resources and Educational Services, or DRES, was established in 1948.

One recent development in this effort, the Speech Accessibility Project, is a collaboration between researchers at the Beckman Institute and corporate partners Microsoft, Amazon, Meta, Google and Apple.

According to the Beckman Institute’s website, the project aims to adapt speech recognition to a more diverse variety of speech patterns, including those with speech disabilities. 

The project will be funded for a minimum of 2 years and will result in an anonymous data set of voice recordings that can be used to train machine learning models and artificial intelligence, according to the Beckman Institute. 

Mark Hasegawa-Johnson, professor in Engineering and Beckman Institute researcher, said that the University’s role in the project is to gather this data and make it available to developers who will then be able to implement it in their own work.

The final products, according to Hasegawa-Johnson, could include speech-to-speech translation, voice-activated smart speakers and smart homes, meeting transcription software and automatic dictation software for writing term papers.

“All five of the (collaborating companies) are committed to using the project data to make their products work better for people with neuromotor speech disabilities,” Hasegawa-Johnson said in a written response. “The project data will also be made available for researchers at other universities and other companies, who may have other ideas.”

Among other reasons, Hasegawa-Johnson believes that the University was selected by partner companies due to its long and impactful history of accessibility research. 

In an email, Mary Bellard, principal innovation architect lead at Microsoft, said the University fit the needs of this project not only because of its commitment to accessibility but also because of the Beckman institute’s unique interdisciplinary approach to research. 

“With programs in computer science, (speech and hearing science), linguistics and accessibility, (the University) not only has a historic commitment to disability inclusion, but the expertise needed to make the project a success,” Bellard said. 

For Microsoft, this project signals movement towards more representation for people with disabilities within their data and products.

“We know one of the biggest challenges the research community and industry at large faces is a (lack of) disability representative data,” Bellard said. “Without diverse, representative data, (machine learning) models cannot learn diversity of speech, limiting customer scenarios where this technology can bring benefits.”

The research team for this project is made up of faculty members from University departments such as linguistics, computer engineering and speech and hearing science.  

One such faculty member, Clarion Mendes, a speech-language pathologist and professor of speech and hearing science, said she believes her team has the potential to make life easier for people with speech disabilities.

“Ideally, researchers will be able to leverage the data to make activities of daily living — and beyond — more attainable for people with disabilities,” Mendes said in a written statement. “The potential for improvements in quality of life and participation are tremendous.”

The success of the project depends not only on the expertise of faculty and student researchers, but also on the input of those who will benefit from the research, she said.

“Our research team is diverse, and includes … Beckman programmers, researchers and students from linguistics, and clinicians and students from Speech & Hearing Science,” Mendes said. “The disability community will be paramount to the success of the project.”

According to Beckman researchers, the findings of this project may begin to impact technology as early as its second year, but they expect most changes to begin after the project is concluded.

[email protected]