Kitchen Waste Classification Based on Deep Residual Network and Transfer Learning

The classification of kitchen waste is of great significance for improving environmental quality and ensuring public health safety. With the development of deep learning technology, automatically classifying garbage has become a hot topic of research. In this paper, a garbage classification algorithm based on deep Residual Network and Transfer Learning is developed, in which the adjusted Residual Network is selected as the base structure. Use the network model trained on ImageNet as a pre-training model, and adopted the model-based Transfer Learning method for learning and optimizing model parameters. Finally, experiment on the garbage classification data set created in this article, and then use AlexNet and VGG-net for comparative experiments. The results show that the algorithm classification accuracy of this paper is up to 86 88% which is higher than AlexNet and VGG-net, respectively 29.64% and 24.37%. This proves the validity of the algorithm in this paper.