Reframing SAR Target Recognition as Visual Reasoning: A Chain-of-Thought Dataset with Multimodal LLMs

TL;DR

This paper introduces a novel approach to SAR target recognition by reformulating it as a multimodal reasoning task using large language models, and presents a new dataset with Chain-of-Thought reasoning to evaluate this method.

Contribution

It pioneers the use of multimodal large language models for SAR target recognition and creates a new dataset with reasoning chains to facilitate this research.

Findings

MLLMs can generate coherent inferences on SAR data

The approach improves interpretability of SAR recognition

Limitations and failure cases of MLLMs are analyzed

Abstract

In the context of Synthetic Aperture Radar (SAR) image recognition, traditional methods often struggle with the intrinsic limitations of SAR data, such as weak texture, high noise, and ambiguous object boundaries. This work explores a novel perspective by reformulating SAR target recognition as a multimodal reasoning task. We leverage multimodal large language models (MLLMs), specifically GPT-4o, to perform target classification based on SAR imagery, guided by candidate categories and enhanced with Chain-of-Thought (CoT) reasoning. A new dataset is constructed based on the FAIR-CSAR benchmark, comprising raw SAR images, structured target annotations, candidate label sets, and GPT-generated CoT reasoning chains. Experimental results show that the MLLMs are capable of generating logically coherent and interpretable inferences in most scenarios. Our analysis highlights both the strengths and current limitations of MLLMs in interpreting SAR imagery, and we provide detailed insights into model behavior through failure case analysis. This work demonstrates the feasibility of incorporating MLLMs into SAR analysis pipelines and establishes a foundation for future research in SAR-oriented visual reasoning.