In this paper, we present a high-performance and light-weight deep learning model for Remote Sensing Image Classification (RSIC), the task of identifying the aerial scene of a remote sensing image. To this end, we first valuate various benchmark convolutional neural network (CNN) architectures: MobileNet V1/V2, ResNet 50/151V2, InceptionV3/InceptionResNetV2, EfficientNet B0/B7, DenseNet 121/201, ConNeXt Tiny/Large. Then, the best performing models are selected to train a compact model in a teacher-student arrangement. The knowledge distillation from the teacher aims to achieve high performance with significantly reduced complexity. By conducting extensive experiments on the NWPU-RESISC45 benchmark, our proposed teacher-student models outperforms the state-of-the-art systems, and has potential to be applied on a wide rage of edge devices.
@article{arxiv.2302.13028,
title = {A Light-weight Deep Learning Model for Remote Sensing Image Classification},
author = {Lam Pham and Cam Le and Dat Ngo and Anh Nguyen and Jasmin Lampert and Alexander Schindler and Ian McLoughlin},
journal= {arXiv preprint arXiv:2302.13028},
year = {2023}
}