SDF-Net: Structure-Aware Disentangled Feature Learning for Opticall-SAR Ship Re-identification

2026-03-12

AI Summary

This work addresses the challenge of cross-modal vessel re-identification between optical and synthetic aperture radar (SAR) images, which arises from their inherent radiometric discrepancies. To tackle this issue, the authors propose a structure-aware decoupled feature learning framework that leverages the geometric stability of vessels as a physical prior. Built upon a Vision Transformer backbone, the method introduces a structural consistency constraint to extract scale-invariant gradient energy statistics. In the final stage, it explicitly disentangles modality-invariant identity features from modality-specific characteristics and enhances discriminability through a parameter-free residual fusion mechanism. Evaluated on the HOSS-ReID dataset, the proposed approach significantly outperforms existing methods, and both code and models have been publicly released.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Cross-modal ship re-identification (ReID) between optical and synthetic aperture radar (SAR) imagery is fundamentally challenged by the severe radiometric discrepancy between passive optical imaging and coherent active radar sensing. While existing approaches primarily rely on statistical distribution alignment or semantic matching, they often overlook a critical physical prior: ships are rigid objects whose geometric structures remain stable across sensing modalities, whereas texture appearance is highly modality-dependent. In this work, we propose SDF-Net, a Structure-Aware Disentangled Feature Learning Network that systematically incorporates geometric consistency into optical--SAR ship ReID. Built upon a ViT backbone, SDF-Net introduces a structure consistency constraint that extracts scale-invariant gradient energy statistics from intermediate layers to robustly anchor representations against radiometric variations. At the terminal stage, SDF-Net disentangles the learned representations into modality-invariant identity features and modality-specific characteristics. These decoupled cues are then integrated through a parameter-free additive residual fusion, effectively enhancing discriminative power. Extensive experiments on the HOSS-ReID dataset demonstrate that SDF-Net consistently outperforms existing state-of-the-art methods. The code and trained models are publicly available at https://github.com/cfrfree/SDF-Net.

Problem

• cross-modal
• ship re-identification
• optical-SAR
• radiometric discrepancy
• geometric structure

Innovation

• Structure-Aware Disentanglement
• Cross-modal Ship Re-identification
• Geometric Consistency
• Modality-Invariant Feature
• Gradient Energy Statistics