ecodetect-waste-classifier

---

EcoDetect Waste Classification Model

This model classifies household waste into 6 categories to help with proper recycling and disposal.

Model Details

• Architecture: EfficientNetB0 with transfer learning
• Accuracy: 91.97% on test set
• Framework: TensorFlow/Keras
• Input Size: 124x124 pixels
• Classes: cardboard, glass, metal, paper, plastic, trash

Usage

import tensorflow as tf
from PIL import Image
import numpy as np

# Load model
model = tf.keras.models.load_model('ecodetect_model.h5')

# Preprocess image
image = Image.open('waste_image.jpg').resize((124, 124))
image_array = np.array(image) / 255.0
image_array = np.expand_dims(image_array, axis=0)

# Predict
predictions = model.predict(image_array)
classes = ['cardboard', 'glass', 'metal', 'paper', 'plastic', 'trash']
predicted_class = classes[np.argmax(predictions)]

Training Data

• Dataset: Trash Type Image Dataset
• Images: 2,527 total images
• Split: 80% train, 10% validation, 10% test
• Data Augmentation: Applied rotation, zoom, flip, and contrast adjustments

Environmental Impact

This model helps users properly classify waste for recycling, contributing to environmental sustainability and waste reduction efforts.