A Robust and Low Complexity Deep Learning Model for Remote Sensing Image Classification
This work addresses efficient image classification for remote sensing applications on edge devices, but it is incremental as it builds on existing low-complexity networks and techniques.
The paper tackled remote sensing image classification by developing a low-complexity deep learning model, achieving competitive performance on the NWPU-RESISC45 dataset with constraints of under 5 million parameters and 20 MB memory.
In this paper, we present a robust and low complexity deep learning model for Remote Sensing Image Classification (RSIC), the task of identifying the scene of a remote sensing image. In particular, we firstly evaluate different low complexity and benchmark deep neural networks: MobileNetV1, MobileNetV2, NASNetMobile, and EfficientNetB0, which present the number of trainable parameters lower than 5 Million (M). After indicating best network architecture, we further improve the network performance by applying attention schemes to multiple feature maps extracted from middle layers of the network. To deal with the issue of increasing the model footprint as using attention schemes, we apply the quantization technique to satisfy the maximum of 20 MB memory occupation. By conducting extensive experiments on the benchmark datasets NWPU-RESISC45, we achieve a robust and low-complexity model, which is very competitive to the state-of-the-art systems and potential for real-life applications on edge devices.