Down syndrome is caused by chromosomal abnormalities and has no link to race, gender or socio-economic status. Their abilities also are varied, but most people with the syndrome do share a common trait — difficulty in being understood.

Most babies born with the syndrome have unique facial features, specifically a small upper jaw and arched palate, along with weak oral muscles, all of which make it hard for them to form words. In addition, children with Down syndrome tend to experience speech delays and often express themselves with key words rather than sentences.

Voice-recognition products are transforming lives, but people with Down syndrome may find themselves left behind. Artificial Intelligence software often doesn't respond to unique forms of speech.

But now, Google has partnered with the Canadian Down Syndrome Society to create Project Understood, an initiative that is helping Google developers create software that recognizes Down syndrome pronunciation and speech patterns. Volunteers are recording ‘scripts’ of sentences that are passed on to Google, giving the tech company the mass of datasets they need for programming. So far, 300 people with Down syndrome have donated their voices to the project — more than half of the goal of 500 recordings.

The end result could be especially beneficial to the more than 250,000 Americans with Down syndrome, including the many adults who are striving to lead independent lives. Being able to voice their needs, order a pizza, call a helper or perform a task through voice recognition will be one big step forward.

If you are interested in volunteering your voice for the project you can contact them here:

Some video imagery courtesy of the Canadian Down Syndrome Society. 

Published on Dec 11, 2019