Garbage Intelligence: Utilizing Vision Transformer for Smart Waste Sorting

Technological advancements have permeated nearly every aspect of modern life, simplifying various human activities. Artificial Intelligence (AI) is pivotal in offering innovative solutions with remarkable efficiency. One crucial area benefiting from AI is garbage or waste management, a fundamental component of contemporary civilization and environmental sustainability. Effective garbage classification is essential for identifying waste types, facilitating appropriate recycling processes, and implementing suitable management strategies. Despite its significance, minimal research has focused on garbage classification, and existing studies often need more regarding accuracy and type of waste. In this research, we propose a novel application of the Vision Transformer (ViT) model for classifying 12 types of household garbage. Our approach leverages deep learning-based image classification capabilities to enhance waste categorization's precision and efficiency. The ViT model performs better than established models such as VGG16 and MobileNetV2. Empirical results show that the ViT model achieves an impressive accuracy of 95%, significantly surpassing the accuracy rates of the other models. This advancement addresses the limitations observed in previous research and contributes to more effective waste management practices by providing a reliable and accurate method for garbage classification. Our study introduces a robust and highly accurate Vision Transformer model for household garbage classification, highlighting its potential to revolutionize waste management systems. By outperforming traditional models, the ViT model sets a new benchmark in the field, paving the way for enhanced recycling processes and environmental sustainability.