Research map / Vision-language and multimodal / Cross-modal learning
Cross-modal Ship Re-Identification via Optical and SAR Imagery: A Novel Dataset and Method
Han Wang, Shengyang Li, Jian Yang, Yuxuan Liu, Yixuan Lv, Zhuang Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add cross-modal dual-head tokenizer (ablation) | ViT-base (no pretraining) · HOSS ReID | mAP: 43.0 -> 44.9 | Improved |
| add dual-head tokenizer after pretraining (ablation) | ViT-base with pretraining · HOSS ReID | mAP: 52.3 -> 53.2 | Improved |
| add contrastive pretraining on optical-SAR pairs (ablation) | ViT-base · HOSS ReID | mAP: 43.0 -> 52.3 | Improved |
| add contrastive pretraining on optical-SAR pairs (ablation) | TransOSS · HOSS ReID | mAP: 49.4 -> 57.4 | Improved |
| add modality information embeddings (ablation) | CDT + pretraining · HOSS ReID | Rank-1: 63.1 -> 64.2 | Improved |
| add ship size embeddings (ablation) | CDT + pretraining · HOSS ReID | mAP: 53.2 -> 55.0 | Improved |
| add both modality and ship size embeddings (ablation) | CDT + pretraining · HOSS ReID | mAP: 53.2 -> 57.4 | Improved |
| use ViT backbones instead of ResNet-based infrared-visible ReID methods | HOSS ReID · HOSS ReID | mAP: ViT-base 43.0 vs CM-NAS 30.7 | Improved |
| use SAR rather than optical as the query modality | TransOSS · HOSS ReID cross-modal | mAP: 48.9 -> 38.7 | Hurt |
| infrared-visible ReID methods applied to optical-SAR | ResNet-based cross-modal ReID methods · HOSS ReID cross-modal | mAP: Optical to SAR mAP 8.2 to 31.3 | No clear effect |
Closest papers by meaning
- SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagery (CVPR 2024 · similarity 0.82)
- Galileo: Learning Global & Local Features of Many Remote Sensing Modalities (ICML 2025 · similarity 0.82)
- Cross-Rejective Open-Set SAR Image Registration (CVPR 2025 · similarity 0.81)
- RobSense: A Robust Multi-modal Foundation Model for Remote Sensing with Static, Temporal, and Incomplete Data Adaptability (CVPR 2025 · similarity 0.81)
- M3LEO: A Multi-Modal, Multi-Label Earth Observation Dataset Integrating Interferometric SAR and Multispectral Data (NeurIPS 2024 · similarity 0.81)
- SARDet-100K: Towards Open-Source Benchmark and ToolKit for Large-Scale SAR Object Detection (NeurIPS 2024 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
