DoNaAI: Deep Learning-Based Ensemble Approach for Waste Classification

Effective solid waste management is crucial in Vietnam due to rapid urbanization. This article presents an AI-based approach to improve waste classification using deep learning models like ResNet, MobileNet, and DenseNet. Trained on ViWaste, a dataset of 10,000 waste images, and enhanced with data augmentation techniques, these models achieved high accuracy in waste classification. Based on these findings, a complete application was developed to support efficient recycling operations. The study demonstrates the potential of deep learning models in promoting environmental sustainability and its broader integration into waste management in developing regions.