Noisy Book

Press

Press is very welcomed, write about Noisy Book and the universe will reward you with gifts.

Maybe.

Press Info.

Noisy Book is a pretty unique App, there isn't anything like it - which is rare.

We spent 2 years working on the idea, but it only became possible with the features of iOS13, which is why Apple invited us to present at WWDC19 this year. You can watch the video here.

The idea for Noisy Book come from a young Mum and her kids. Luckily enough, Dad was an App Developer, so this is a lovely family venture - and a proper Indie Dev story.

Download The Press Pack here. Screenhots, GIFs, Logos, all included and ready to use.

Not Technical Stuff.

The idea of Noisy Book was to empower young kids to enjoy reading. We wanted them to pick up a physical book and want to read it. Noisy Book does this by providing automatic sound effects based on your speech, and it also does some nice visual effects too. It's a really magical experience.

Along the journey of creating the app, it became evident that Noisy Book was so powerful, it was also a brilliant creative tool for older kids, for them to explore their creative story telling, and make up their own stories adlib to the app. It was amazing to watch kids just use the app, and their own imaginations, so we have also focused our attention to this area.

Technical Stuff.

Noisy Book looks simple on the outside, but there is a TON of stuff happening on the inside! Part of our magic, is making something complex seem so simple. Before iOS13, this app wasn't even possible (believe us, we tried!). Thanks to the new CoreML and Natural Language Processing features in iOS13, we were finally able to make our dream come alive.

Here's a headline list of the major technology in the app:

  • Supports 7 languages
  • On Device Speech Recognition
  • Natural Language Processing, Apple API
  • Natural Language Processing, custom CoreML model
  • Speech Saliency, iOS 13
  • Book and Part of book recognition, custom CoreML model
  • Sound Analysis, iOS13, custom CoreML model
  • Word Vectors and Sound Vectors for sound effect matching

With all of this technology, we are able to understand not just what is said, but also how it is said. This lets us create a beautiful transcription of a story based on speech inflection and saliency.

We use the what was said to join together the actors, objects, animals and scenes in the story, then use that to create an end-of-story Quiz, which is dynamically generated. This follows the young education area of story recall, and can ask the readers about their story: what happened, who did it happen too, or even - ask them to make the sounds they heard back to the app.

Contact Us