PCM-SAR: Physics-Driven Contrastive Mutual Learning for SAR Classification

2025-04-13

AI Summary

Existing SAR image classification methods largely adopt contrastive learning strategies designed for optical imagery, neglecting SAR-specific physical characteristics—such as speckle noise and geometric distortions—leading to suboptimal feature representation. To address this, we propose a physics-driven contrastive mutual learning framework. First, we introduce a novel contrastive learning paradigm explicitly incorporating SAR imaging physics priors. Second, we design an unsupervised local sampling strategy guided by semantic detection and speckle noise simulation modeled via Gray-Level Co-occurrence Matrix (GLCM), enabling domain-adaptive sample generation. Third, we integrate multi-level feature fusion with a teacher-student mutual learning mechanism to enhance the discriminative capability of lightweight models. Extensive experiments on multiple benchmark SAR datasets demonstrate significant improvements over state-of-the-art methods: our lightweight model achieves an average classification accuracy gain of 4.2%, empirically validating the critical role of physics-guided supervision in SAR representation learning.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Low Level & Physics-based Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Existing SAR image classification methods based on Contrastive Learning often rely on sample generation strategies designed for optical images, failing to capture the distinct semantic and physical characteristics of SAR data. To address this, we propose Physics-Driven Contrastive Mutual Learning for SAR Classification (PCM-SAR), which incorporates domain-specific physical insights to improve sample generation and feature extraction. PCM-SAR utilizes the gray-level co-occurrence matrix (GLCM) to simulate realistic noise patterns and applies semantic detection for unsupervised local sampling, ensuring generated samples accurately reflect SAR imaging properties. Additionally, a multi-level feature fusion mechanism based on mutual learning enables collaborative refinement of feature representations. Notably, PCM-SAR significantly enhances smaller models by refining SAR feature representations, compensating for their limited capacity. Experimental results show that PCM-SAR consistently outperforms SOTA methods across diverse datasets and SAR classification tasks.

Problem

• Improving SAR classification by capturing unique semantic and physical characteristics
• Enhancing sample generation with domain-specific physical insights and GLCM
• Boosting smaller models via refined feature representations and mutual learning

Innovation

• Uses GLCM for realistic SAR noise simulation
• Applies semantic detection for unsupervised sampling
• Multi-level feature fusion via mutual learning