An EU project to support local radio has created a synthesised Irish-speaking voice that will take over segments of a Cork station.
As part of the EU’s Grassroot Wavelengths initiative, a company called CereProc has produced what it says is the first commercially available Irish language synthetic voice, which will soon take to the airwaves.
With participants across Portugal, Romania and Ireland, the EU initiative aims to create a network of inclusive digital platforms, enabling the free flow of information within geographic communities. Now, this text-to-speech technology will debut on a community station on Bere Island in south-west Cork.
When the station can’t be staffed by any human presenters, the digital DJ will generate spoken content in both English with an Irish accent, and the Irish language. Despite it first launching in Cork, the voice is based on a regional Donegal accent.
The voice technology being used for this radio station has been made available on multiple platforms including Microsoft, Android and iOS, and is for personal, business or commercial use.
While there is a fee for business and commercial access, it has been made available free of charge for personal use to help anyone with communication difficulties to speak digitally with a more local accent.
Dr Matthew Aylett, chief scientific officer at CereProc, said that there is a demand for different accents from synthetic voices because “Alexa or Siri are dull and all very familiar”.
“It’s time for technology to support diversity rather than encourage conformity,” he said.
“The Irish language is key to Irish identity and our voice can be used in tandem with our Irish-accented English language voice to create a more personalised experience for the Irish market and beyond.”
He added that such initiatives are needed to break the hold major corporations have over voice technologies.
“With just a few companies threatening to dominate the bulk of our technological and online experience, it’s imperative that we embrace difference and develop technology that can be used by all,” Aylett said.
“We are proud to champion less widely used languages and are committed to developing affordable synthesised voices that can accommodate regional accents from across the globe.”