AI-Powered Trash Classification System for Lakes and Water Bodies Using Transfer Learning

Abstract

The growing accumulation of rubbish in lakes and bodies of water poses a huge environmental concern that necessitates creative solutions. To address this issue, the study proposes an artificial intelligence (AI) powered trash classification system based on transfer learning. This system seeks to revolutionize waste management in lakes and water bodies by leveraging AI and machine learning (ML) technology. The methodology integrates computer vision, advanced ML algorithms, and robotics to identify, classify, and collect various sorts of trash in bodies of water. The system incorporates a pre-trained residual network (ResN et-50) model. Extensive simulations are carried out, and the experiments show that transfer learning using ResN et50 can obtain great results in classifying garbage photos into numerous categories. With an accuracy of 95.6%, the model offers a promising deep-learning option for automated waste management systems. The model's strong precision and recall scores demonstrate its ability to classify various waste types in real-world contexts correctly.