Transfer Learning on EfficientNet for Remote Sensing image Classification

Nowadays remote sensing image classification plays an important role in various applications and has become an active research topic. A variety of approaches have been proposed for image classification, especially based on convolutional neural networks(CNN). However, although many methods have achieved good accuracy, their network structure is very large and the training parameters are considerable, which affect the overall performance. In this paper, we present a transfer learning method based on pre-trained EfficientNet models with fine tuning strategy for remote sensing image classification. EfficientNet achieves the current state-of-the-art performance using significantly less parameters than other latest models in this field of image classification. But nowadays the network is still lacking used in remote sensing tasks. Our method is validated on five remote sensing data sets, the experimental results show the effectiveness and superiority of the proposed methods for scene classification in remote sensing imagery.