Waste classification has attracted more and more attention in recent years, which is an important part of building an eco-friendly city. Traditional manual garbage classification has poor efficiency and accuracy. In this paper, based on deep learning, the garbage classification algorithm I-ResNet50 is proposed to improve the ResNet50 network, and the geometric transformation of the original data is performed. The test set results show that the I-ResNet50 algorithm can achieve a classification accuracy of 62.6%, which is a substantial improvement in accuracy compared with the original method.
Paper
Full text
Deep learning based recyclable waste classification
Semantic Scholar · Environmental Science · 2022
Abstract
Waste classification has attracted more and more attention in recent years, which is an important part of building an eco-friendly city. Traditional manual garbage classification has poor efficiency and accuracy. In this paper, based on deep learning, the garbage classification algorithm I-ResNet50 is proposed to improve the ResNet50 network, and the geometric transformation of the original data is performed. The test set results show that the I-ResNet50 algorithm can achieve a classification accuracy of 62.6%, which is a substantial improvement in accuracy compared with the original method.