Waste_classification_Machine_Learning

Explored transfer learning with various pretrained models on waste classification dataset Evaluated the usage semi‐supervised learning using pseudo labelling. With the best model based on accuracy and precision, built a Fast API web server. Created a website in Svelte, upon capturing an image, made calls to the FastAPI to make predictions in the frontend. Deployed the trained model to the google cloud platform and used it instead for the backend API calls.

If you want to check out my work: https://waste-class-frontend-svelte.vercel.app/