Smart Waste Segregation and Composting using AI and IoT

This work presents a smart waste segregation and composting framework integrating Artificial Intelligence (AI) and Internet of Things (IoT) for automated, efficient, and sustainable solid-waste management.The system employs YOLOv5 for static uploaded images and YOLOv8 for live image to accurately detect waste items across varying real-time conditions.Detected segments are further classified using a hybrid ResNet-50 and Transformer model, leveraging TorchVision's feature extraction and attention-based refinement.A Flask server manages inference workflows and communicates with IoT-based segregation and composting units.Sensors inside the compost chamber continuously monitor temperature, humidity, and gas emission levels to maintain optimal decomposition.Experimental evaluation demonstrates high detection precision, improved classification accuracy, low processing latency, and enhanced compost quality.The proposed system offers an automated, scalable, and eco-friendly approach for modern smart cities.