Visualization of Deep Transfer Learning In SAR Imagery

TL;DR

This paper explores transfer learning from electro-optical to SAR imagery, visualizing how deep networks interpret SAR data using class-activation maps to understand feature transfer.

Contribution

It demonstrates the application of transfer learning from EO datasets to SAR imagery and visualizes the process with class-activation maps for interpretability.

Findings

Transfer learning enables SAR image analysis using EO-trained models.

Class-activation maps reveal how features are transferred and interpreted.

Visualization provides insights into deep network behavior on SAR data.

Abstract

Synthetic Aperture Radar (SAR) imagery has diverse applications in land and marine surveillance. Unlike electro-optical (EO) systems, these systems are not affected by weather conditions and can be used in the day and night times. With the growing importance of SAR imagery, it would be desirable if models trained on widely available EO datasets can also be used for SAR images. In this work, we consider transfer learning to leverage deep features from a network trained on an EO ships dataset and generate predictions on SAR imagery. Furthermore, by exploring the network activations in the form of class-activation maps (CAMs), we visualize the transfer learning process to SAR imagery and gain insight on how a deep network interprets a new modality.