Speaker:
Tero Parviainen

Abstract:
Imagine the sound of a forest meadow. Or a crackling fireplace. Or a street corner in your favourite city. Or the bottom of the sea. Or the cargo hold of a drifting interstellar spaceship. What does it sound like there?

Well, thanks to the increasingly strange technologies we now have in our hands, we can now just go to these places and listen to them – kind of! We can generate machine portrayals of all kinds of places that exist, as well as places that definitely don’t.

In this talk I’ll give a technical runthrough of how an immersive sonic experience like this can be built with a combination of web technologies, generative audio models, foundation AI models, speech-to-text, prompt engineering, and a bit of magic.

—

Silver sponsors: Smartly, Wunderdog, Alma Media, Backscreen
Bronzer sponsor: Wonna

Future Frontend 2025 took place 27-28.5.25 at Dipoli, Espoo, Finland. You can learn more about the conference at https://futurefrontend.com.

source