eco-waste-detector

• 

• 

  https://image2url.com/r2/default/videos/1774047004153-68713c5f-744f-43f1-8ac0-2f8402fb62c6.mp4

Eco Waste Detector was inspired by something I’ve always wondered about instead of people having to manually sort all the waste, what if machines could recognize it and do it for us? I wanted to build something simple but actually useful that could guide people in real time using AI. While working on this project, I learned how image classification works and how to train models using Teachable Machine. I also improved my skills in JavaScript and HTML, especially when it came to connecting a machine learning model to a web application using TensorFlow.js. I built the project by first training a model to recognize different types of waste like recyclable, organic, and landfill items. After training, I exported the model and integrated it into a web interface. The website uses the webcam to capture images and the model predicts the type of waste. One of the biggest challenges I faced was getting accurate predictions, since lighting, angles, and different objects can confuse the model. I also had issues connecting the model properly to the frontend. Debugging these problems took time, but it helped me better understand how everything works together. Overall, this project was a great learning experience, and I still want to keep improving it by making the model more accurate and adding more categories. At the same time, I’m hoping that in the future machines like this can actually handle waste sorting completely, so everything can be sorted faster and more effectively.

Built With

• html
• javascript
• teachable-machine
• tensorflow.js
• web