Improvement of EfficientNet in medical waste classification

Abstract

In recent years, medical waste has gained attention due to its hazardous nature, complexity, and high cost of manual sorting and management. Therefore, it is crucial to develop classification systems that are accurate and efficient. This study analyzes various deep learning models for medical waste classification, compares their accuracies in image recognition, and provides an in-depth analysis of EfficientNet, a classification model that is well-suited to handle large amounts of waste mixing. EfficientNet’s superior performance can be adapted to numerous potential scenarios in medical waste, and its improved performance is also very promising in the field of medical waste classification. The data demonstrate its significant advantages over other models, indicating broad application prospects and economic benefits.