Inspiration

  • Help the environment by reducing waste that goes into landfill

What it does

  • Closes the gap in information for recycling. Any user can hold any item they want to recycle and be informed of how they should recycle it

How we built it

  • We used tensorflow.js a machine learning API to build an algorithm and then trained it on images to differentiate between "Recycle", "Compost", and "Trash"

Challenges we ran into

  • Originally, when we were trying to localhost the website, we ran into a lot of permissions issues. A lot of our other API's such as Node.JS were giving us trouble.

Accomplishments that we're proud of

  • Working tensorflow algorithm
  • Beautiful website to compliment it all

What we learned

  • Overall better usage of HTML & CSS
  • API knowledge grew significantly
  • Learned about localhost such as XAAMP

What's next for EcoSorter

  • Given more time, we can continue expanding the ML algorithm. Our website is POC where we wanted to prove
  • We look forward to implement code that will let the algorithm detect the item's category without needing to press any buttons to make it memorize

Built With

Share this project:

Updates