Ai-like

Focused on helping elementary aged children understand artificial intelligence through low level explanations of how AI works and activities involving image identification processes similar to AI.

Comment

Inspiration

AI is currently in its advent. It is important to know about the intuition of AI as elementary school children. which is really necessary as they have a lot of economical and social impact on our lives. So, understanding AI is necessary as the children are the pillars of the next generation that will lead the society so they must have an understanding of the mechanics and the ethical issues behind using AI and financial implications. So, we present to you Ai-Like.

What it does

Ai-Like is an online AI education platform that introduces some advanced concepts of AI through simple intuitions. We are aiming to pique interest in elementary school children to understand AI and show them a general overview of what it does. Our current submission focuses on two modules - What is AI and introducing CNNs through simple intuitions to the targeted age group. Our Vision is to extend this work to include a learning path and also for different sub-age groups within elementary school children.

How we built it

We Built this platform using React.js and Flask. The training modules' content has been handcrafted using TensorFlow APIs where we show some interesting results in our app to the children. We have also created an interactive app that will help the children to get hands-on experience on CNNs.

Challenges we ran into

  • We ran into problems getting Google OAuth integrated. These issues were due to some security settings with respect to Google OAuth.

Accomplishments that we're proud of

  • Integrating multiple components of front end and backend using flask and React.js
  • We were also able to integrate the flask mail and Twilio API for feedback about students.
  • We were also able to create content that will help children get an intuition of what AI is by breaking it down in the highest form of abstraction for the targeted age group.
  • We were also able to showcase our expertise in AI by creating code that will generate guided backdrop views and saliency maps for the CNNs that we have trained.
  • We were also able to create a flask app that could provide a hands-on tool for children that was trained on ImageNet(80 classes) to demonstrate what CNNs could do.

What we learned

-Integration of frontend and backend components with respect to React.js and Flask

-Bootstrap Material theme components usage

-Viewing Filters activations and Saliency maps to create explainable AI models

-Create a Flask app for CNNs that could do inferencing in realtime.

  • Creating content for targeted age groups.

  • Working remotely through version control(GitHub)

  • Scrum methodologies

  • Project Management

  • Software Engineering

What's next for Ai-Like

  • The immediate goals for AI-Like are to create content at different levels for the targeted audience age group(Elementary School).
  • The future vision is to scale up with multiple interactive features and translation APIs

Built With

Share this project:

Updates