Waste Classifier

• Tech Stack: TensorFlow, OpenCV, NumPy, Matplotlib, GoogleColab,Flask, HTML,CSS,JavaScript.
• Github URL: Github Link

• Developed a Convolutional Neural Network (CNN) to classify waste materials into 8 categories: cardboard, plastic, metal, glass, food waste, electronics, paper, and trash.

• Trained the model on 1,200 TrashNet dataset samples, achieving 79% accuracy on test data and 91% accuracy on training data.

• Optimized the CNN model using hyperparameter tuning and regularizations (e.g., dropout, L2 regularization), improving model performance and reducing overfitting.