Semantic Smoothing via Novel View Synthesis for Robust SAR Image Classification

TL;DR

This paper introduces semantic smoothing, a novel adversarial defense for SAR image classification that uses structured view synthesis to generate multiple plausible radar views, enhancing robustness and accuracy.

Contribution

The paper proposes a new semantic smoothing method that employs view synthesis conditioned on acquisition geometry to improve adversarial robustness in SAR ATR.

Findings

Semantic smoothing enhances robustness against FGSM, PGD, OTSA, and SMGAA attacks.

It increases clean classification accuracy.

Structured transformations outperform isotropic noise in preserving SAR image semantics.

Abstract

Deep neural networks are vulnerable to adversarial perturbations, limiting deployment in safety-critical applications such as synthetic aperture radar (SAR) automatic target recognition (ATR). Randomized smoothing improves robustness by averaging predictions over noisy inputs, but isotropic noise often fails to preserve the semantic structure of SAR imagery. We propose semantic smoothing, a defense that replaces noised-based perturbations with structured randomized transformations generated by a novel view synthesis model. For SAR, we condition on acquisition geometry to synthesize multiple plausible radar views. Predictions across generated randomized views are aggregated to form a robust classifier. Experiments show that semantic smoothing improves robustness against standard attacks, such as FGSM and PGD, and SAR-specific attacks, such as OTSA and SMGAA, while also increasing clean classification accuracy. These results demonstrate that randomized smoothing via semantically preserving geometric transformations is a promising alternative to isotropic noise for adversarial defense in structured sensing domains.