From text to audio. It’s a technology that’s been around for years. However, Leiden startup Daisys believes it has developed something new: a computer voice that is generated using artificial intelligence (AI), while the popular reading technology relies on audio parts of people. The AI sound looks natural and can be adjusted in real time in terms of speed and tone.
According to Daisys, the 17th century philosopher Descartes was wrong when he said “I think, therefore I am.” The phrase should be “I look, therefore I am.” And so they gave British computer pioneer Alan Turing posthumous fame.
Daisys generates realistic computer sounds that never existed before. Deepfake technology has not been used for existing audio portions, as is typical with text-to-speech technology until now. According to director Barnier Geerling, such deepfakes are untenable. After all, not everyone wants to borrow without controlling what is being said.
train models
“Descartes was wrong, and Turing was right”
For the past year and a half, the startup has been working on the artificial intelligence technology behind Reading Voices. “We have provided the existing core technology with some significant modifications,” says chief technology officer Joost Broekens. “In addition, we had to ‘train’ our models in a smart way, using the correct ratio of speech data from different speakers.”
The fact that the reading sound is entirely generated by an algorithm means that it is easy to tune and yet sounds natural. According to the company, this technology can be used anywhere a human voice is needed. This concerns, for example, media, smart devices, games, robots, speech aids, and public speech systems.
employment Location Daisys has an audio portion of the technology. The sounds you actually hear sound more natural than many other computer-generated reading sounds. There are even some recognizable feelings, although it’s still obvious that they are artificial sounds. “If everyone thinks I’m real, then I’m real (…) so Descartes was wrong and Turing was right. I seem, therefore I am,” said the voices.
