Artificial IntelliKid

Page 1
Page 2
Page 3
Page 4

Inspiration

We made a website that introduces children to deep learning, AI, and IoT. We were inspired by an introduction to bitcoin website we've seen before with slides, explanations, and demos.

What it does

It explains deep learning from a very high level and then gets into some details about the math behind it without overwhelming kids. In the end it opens their minds to all the possibilities of the deep learning breakthrough.

How we built it

We built it using React and Nextjs for UI, tensorflowjs and mlrjs for the AI.

Challenges we ran into

We ran out of time and would have liked to include a demo of Shazam or another audio analyzer as well as a visual analyzer but these wouldn't be too hard to procure with some extra time.

Accomplishments that we're proud of

We got some pre-trained models to work with react and browser.

What we learned

Ai is hard in browser, and the models are very particular about input sizes.

What's next for AI Challenge

This challenge needs integration with the established Shazam API as well as fancier visual demos, such as facial recognition.

Built With

javascript
ml5.js
next.js
react
tensorflow

Submitted to

HackGSU 2022
- Winner GSU CS Department Challenge #1 - 1st Place

Created by

I'm responsible for the front-end development. I work close with the AI models resource to present an interactive and fun experience for the user. This is also my first time using Next.js and it was such a fun time working with this framework. I also responsible for the project's brand.

Cuong Hoang
I did the writing and some of the html. I described the limits, capabilities, and processes of machine learning by using analogies and examples for kids.

alphacentauri718 Kanwar
I was responsible for the ideation of the project and embedding preexisting AI models using ml5.js and a trained Tensorflow.js model into the project pages.

Dileep Reddy Pemmana
Gia Hiển Trần