Boundary-Aligned Contribution Routing for Robust Optical--SAR Object Detection

2026-08-15

AI Summary

This study addresses the negative transfer issue arising from imperfect modal correspondence in optical-SAR fusion detection by proposing a Boundary-Aligned Contribution Routing mechanism. Leveraging feature-level and dual-statistic semantic routers, this method dynamically modulates modal contributions prior to feature mixing without requiring additional utility supervision, thereby effectively suppressing negative transfer. Experimental results demonstrate that the proposed mechanism significantly enhances fusion robustness, improving mean Average Precision (mAP) by 0.5–5.9 points under full-input conditions and 7.6–41.6 points in missing-modality scenarios. Furthermore, it reduces the negative transfer rate by up to 12.7%, successfully mitigating performance degradation challenges inherent in heterogeneous multimodal fusion tasks.

Abstract

Optical imagery provides rich appearance cues, whereas synthetic aperture radar (SAR) offers observations that are less sensitive to illumination and weather, making optical--SAR fusion attractive for remote-sensing object detection. However, the presence of multiple modalities does not guarantee beneficial fusion: imperfect spatial, temporal, and semantic correspondence can make an otherwise intact stream conditionally harmful and induce negative cross-modal transfer. We handle this issue through a model-specific task-utility perspective and learn task-conditioned contribution routing using detection supervision alone. The proposed fusion-boundary-aligned routing regulates each modality's contribution before the first learned cross-modal feature-value mixing operation. For architectures with frequent shallow interaction, a Feature Router performs cross-conditioned, group-addressable modulation near the input; for dual-backbone architectures, a Dual-Statistic Semantic Router predicts stream-level contribution weights from modality-specific average and maximum statistics before late semantic fusion. The routers require no explicit utility supervision, quality labels, reconstruction, or distillation. Experiments on M4-SAR and SpaceNet6-OTD cover nominal full inputs, controlled correspondence shifts, missing modalities, and four nonzero modality-corruption scenarios. Across the reported clean-training controls, routing improves full-input $\text{mAP}_{50}$ by 0.5--5.9 points. Relative to the corresponding modality-dropout baselines, it raises missing-modality $\text{mAP}_{50}$ by 7.6--41.6 points and reduces the negative-transfer rate by up to 12.7 percentage points. Spearman correlations between the learned routing weights and model-specific leave-one-modality-out utility range from 0.45 to 0.66, supporting the task-utility interpretation of the routing coefficients.

Problem

• Optical-SAR fusion
• Negative cross-modal transfer
• Object detection
• Imperfect correspondence

Innovation

• Contribution Routing
• Negative Transfer Mitigation
• Task-Conditioned Modulation
• Optical-SAR Fusion
• Boundary-Aligned Routing