Big tech companies like Microsoft, Apple, Google, Meta and Amazon are working to develop speech recognition technologies for people with disabilities. These companies are collaborating with the University of Illinois (UIUC) as part of the Accessibility of Broadcasting project.

The project aims to improve speech recognition for users with ALS, more commonly known as Lou Gehrig’s disease, cerebral palsy, Down’s syndrome or Parkinson’s disease, who have altered speech patterns. Microsoft has recently improved Teams calls on Windows with artificial intelligence, and Amazon has launched new Echo devices with a better voice architecture, such as the Echo Dot and Alexa Voice Remote Pro. Perhaps the project can help popularize such products for people with different speech patterns in the future.

Nonprofits such as the Davis Finney Parkinson Foundation and Team Gleason have also joined the project. It focuses on how, despite advances in voice recognition technology and translation tools, it is a barrier for people with different speech patterns.

UIUC professor Mark Hasegawa-Johnson said:

“Language interfaces must be accessible to everyone, including people with disabilities. This challenge was challenging because it required a large amount of infrastructure that could ideally be supported by leading tech companies, so we are a unique cross-disciplinary team with expertise in linguistics, speech, artificial intelligence, security and privacy. We built a team.”

The initiative will take speech samples from users with different speech patterns or speech disabilities to build a dataset to train machine learning models. For this, paid volunteers will be recruited initially in American English. Source