Multi-class Waste Segregation using EfficientNetb3 Model through Waste Segregation Dataset

Efficient waste segregation is critical for successful recycling and sustainable waste management methods. In this paper, we investigate the use of EfficientNetB3, a cutting-edge deep learning model, to obtain high accuracy in trash segregation tasks. Using the Adam optimizer and training the model for 30 epochs, achieving 97% accuracy rate in categorising various waste types. The Adam optimizer's adaptive learning rate approach improves model efficiency and convergence, allowing for rapid training and durable performance. Deep learning algorithms, particularly EfficientNetB3, have the potential to automate trash segregation procedures, streamline recycling efforts, and enhance environmental sustainability. This study helps to advance intelligent waste management systems by demonstrating the efficacy of AI-driven approaches in addressing difficult environmental concerns. Future research will concentrate on incorporating AI technology into practical problems waste management infrastructures to improve efficiency and support circular economy concepts.