Intelligent Waste Material Classification Using EfficientNet-B3 Convolutional Neural Network for Enhanced Waste Management

In urban environments, the escalating accumulation of solid waste has emerged as a pressing environmental and health concern, underscoring the need for effective waste management strategies. A critical facet of this process involves segregating waste into distinct components, traditionally achieved through manual sorting methods. To streamline and enhance this procedure, this study proposes an intelligent waste material classification system. Leveraging the capabilities of the EfficientNet-B3 Convolutional Neural Network-a pre-trained model with 28 layers-this study aims to categorize waste images into eight specific groups, encompassing materials like glass, metal, paper, plastic, and more. Through rigorous evaluation on a diverse waste image dataset, proposed system achieves an impressive 97% accuracy. The deployment of this waste material classification system holds the potential to expedite and optimize waste separation, thereby reducing the dependence on extensive human intervention. This research contributes to the advancement of intelligent waste management practices, offering a promising avenue for sustainable urban development.