Published: 23 Jun 2022 16:25 GMT
Amazon’s senior vice president and chief scientist for Alexa, Rohit Prasad, presented this Wednesday at the annual re:Mars conference in Las Vegas, a potential feature of the company’s digital assistant, which consists of the ability to revive the voice of any person from the synthesis of a short audio recording.
In this way, the company aims to bring back to users the voices of their deceased loved ones. During the presentation, Alexa read part of the story ‘The Wizard of Oz’ to a child in the voice of his deceased grandmother.
Prasad noted that Amazon is now able to achieve impressive results relying on just one minute of recorded voice. “This required inventions where we had to learn how to produce a high-quality voice with less than a minute of recording, versus hours of recording in the studio,” he said.
“The way we achieved this was by framing the problem as a speech conversion task rather than a speech generation path. There is no doubt that we are living in the golden age of artificial intelligence, where our scientific dreams and fictions are becoming reality,” said the senior executive.
At the moment, not many details are known about Alexa’s new capabilities, nor have the timeframes in which they might be available to users transcended.