At the company's annual re:Mars conference in Las Vegas, Amazon's senior vice president and head scientist for the company's smart assistant announced a number of new features. There was a new feature that could make short audio clips into longer speech.
In the scenario presented at the event, the voice of a dead loved one is used to read a grandson a book. The company is able to achieve some very impressive audio output using just one minute of speech.
In order to make a high-quality voice with less than a minute of recording, we had to invent inventions. We framed the problem as a voice conversion task and not a speech generation path. The golden era of artificial intelligence is where our dreams and science are.
There are very little details at the moment. This is the kind of news that will likely invite all manner of scrutiny over potential applications beyond just reading a child The Wizard of Oz.