EcoDispose

An AI-powered waste classifier and disposal method generating software

Comment

Inspiration

My inspiration was noticing how much waste goes into trash cans and bins and they are taken to landfill sites. I felt it would be easier if people had something to classify waste and suggest waste disposal methods for them.

What it does

EcoDispose accepts image of waste, and classifies them into Organic or Recyclable. Once it has classified the waste, it suggests disposal methods for said waste type.

How we built it

I built the classifier using convolutional neural networks and i utilized ChatGPT 4o API to suggest disposal methods. Flask was used for the backend, and HTML,CSS and Javascript were used to make the frontend.

Challenges we ran into

Some challenges included fine-tuning and training the model to overcome image outliers, and making the user interface run smoothly and appear modern at the same time. But overall, with a bit of effort, I was able to overcome said challenges.

Accomplishments that we're proud of

The EcoDispose platform utilises deep learning for the image analysis, and the frontend was made using HTML-CSS-JS. I am proud of this since I learned deep learning and was able to integrate it with a fairly simple frontend service like HTML.

What we learned

I learned how to implement deep learning models and connect with Flask apps and carry out image recognition. I also learned how to implement a simple, yet modern user interface for EcoDispose by using HTML-CS-JS which uses less system resources and load time than ReactJS, Next.js and others.

What's next for EcoDispose

In the future, we could expand the platform. Rather than being a simple waste classifier, it will also allow users to sign up, classify waste, earn points and gain rewards the more they recycle. This will encourage people into participating in activities that help the environment and take steps for a better, greener future.

Built With

Share this project:

Updates