A Task-Driven and Quality-Assured Agent Framework for SAR Data Generation

2026-06-27

AI Summary

This work addresses the challenges in synthetic aperture radar (SAR) data augmentation, including heterogeneous data formats, strong task dependency, diverse generation methodologies, and weak validation mechanisms, which collectively hinder reproducibility and limit support for downstream tasks. To overcome these issues, the authors propose SAGA, a novel framework that decouples semantic proposal from deterministic verification for the first time, enabling a task-oriented, quality-controlled, and auditable SAR data augmentation agent. SAGA integrates natural language understanding, constraint-based planning, and workflow compilation, and introduces a multidimensional evaluation protocol encompassing data quality, distribution fidelity, SAR-specific artifacts, redundancy, data leakage, and downstream performance. Experiments demonstrate that SAGA significantly outperforms baseline approaches—including rule-based methods, pure large language models, ReAct, and fixed augmentation strategies—in pattern alignment, strategy planning, rejection of invalid samples, and overall augmentation effectiveness.

Abstract

Synthetic aperture radar (SAR) data augmentation is important for improving the generalization of data-driven SAR interpretation models, yet practical augmentation workflows are often hindered by heterogeneous dataset formats, task-dependent metadata requirements, diverse generation methods, and weak validation of generated samples. This paper presents the \textbf{S}AR \textbf{A}ugmentation and \textbf{G}eneration \textbf{A}gent (SAGA), a schema-grounded and benefit-aware agent framework for task-oriented SAR data generation and augmentation. Given a natural-language request and heterogeneous SAR inputs, SAGA extracts observable dataset facts, validates executable dataset schemas, selects feasible augmentation strategies through validator-constrained planning, and compiles the selected strategy into an auditable augmentation workflow. Generated data are further assessed by quality, distribution, SAR-artifact, duplicate, leakage, and optional downstream-task evaluators to support evidence-qualified augmentation claims. By separating semantic proposal from deterministic validation and execution, SAGA improves the reliability and reproducibility of SAR augmentation decisions. Experiments on controlled agentic benchmarks and downstream SAR interpretation tasks show that SAGA improves schema grounding, skill planning, invalid-sample rejection, and downstream augmentation utility compared with rule-based, LLM-only, ReAct-style, and fixed-augmentation baselines.

Problem

• SAR data augmentation
• heterogeneous dataset formats
• task-dependent metadata
• generation validation
• data-driven interpretation

Innovation

• SAR data augmentation
• agent framework
• schema grounding
• quality assurance
• task-driven generation