If you ever felt that Alexa should be able to read in a newscaster’s voice while giving you your daily flash briefing, the voice-activated assistant will soon be able to do that.
Amazon stated in a recent blogpost that its neural text-to-speech system “can learn to employ a newscaster style from just a few hours of training data.” The system utilizes a generative neural network that uses machine learning to generate voices better.
Listen to what the newscaster might sound like here.
For users listening to the speech, the language is more natural than a sound produced through “concatenative methods, which string together short speech snippets stored in an audio database,” according to Amazon.
“This advance paves the way for Alexa and other services to adopt different speaking styles in different contexts, improving customer experiences,” the blogpost said.
It’s likely that Amazon has plans to release more Alexa voices throughout 2019.