Rohit Prasad, Amazon's head scientist for Alexa, said at a live event on Wednesday that his team has been instructing Alexa to pick up a voice from a short audio clip and convert it into longer audio output. Prasad was presenting at Amazon's re:Mars conference in Las Vegas.
He showed a short video of how people could use Alexa's voice-changing capability in real life. In the clip, a boy asks: "Alexa, can grandma finish reading me the Wizard of Oz?"
The smart speaker confirmed the request in its default chirpy voice, then transitioned to a less robotic voice that narrated an excerpt from the children's novel.
"This required inventions where we had to learn to produce a high-quality voice with less than a minute of recording versus hours of recording in the studio. The way we made it happen is by framing the problem as a voice-conversion task and not a speech generation path," Prasad said.
Prasad said Alexa's ability to impersonate familiar voices is particularly crucial now, as many people lost loved ones to COVID-19.