Research map / Detection and segmentation / 3D detection and driving
M4-SAR: A Multi-Resolution, Multi-Polarization, Multi-Scene, Multi-Source Dataset and Benchmark for optical-SAR Object Detection
Chao Wang, Wei Lu, Xiang Li, Jian Yang, Lei Luo
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 FAM to baseline (ablation) | CSSA baseline · M4-SAR | mAP: 58.0 -> 58.9 | Improved |
| add CMIM to baseline (ablation) | CSSA baseline · M4-SAR | mAP: 58.0 -> 59.6 | Improved |
| add AFM to baseline (ablation) | CSSA baseline · M4-SAR | mAP: 58.0 -> 60.0 | Improved |
| add AFM to FAM+CMIM (ablation) | E2E-OSDet · M4-SAR | mAP: 61.0 -> 61.4 | Improved |
| add CMIM to FAM+AFM (ablation) | E2E-OSDet · M4-SAR | mAP: 60.7 -> 61.4 | Improved |
| add FAM to CMIM+AFM (ablation) | E2E-OSDet · M4-SAR | mAP: 60.9 -> 61.4 | Improved |
| fuse optical with SAR instead of optical only | E2E-OSDet vs YOLO-Master · M4-SAR | mAP: 56.3 -> 61.4 | Improved |
| add SAR to optical for fusion detection vs SAR-only | E2E-OSDet vs YOLO-Master SAR · M4-SAR | mAP: 48.5 -> 61.4 | Improved |
| mix optical and SAR data to train a single detector | YOLO26 · M4-SAR | mAP: 55.4 -> 47.3 | Hurt |
| mix optical and SAR data to train a single detector | YOLO-Master · M4-SAR | mAP: 56.3 -> 49.0 | Hurt |
| use Grad handcrafted feature in FAM versus other descriptors (ablation) | FAM · M4-SAR | mAP | Improved |
| vary feature proportion and area size (ablation) | E2E-OSDet · M4-SAR | mAP | No clear effect |
Closest papers by meaning
- SARDet-100K: Towards Open-Source Benchmark and ToolKit for Large-Scale SAR Object Detection (NeurIPS 2024 · similarity 0.89)
- RSAR: Restricted State Angle Resolver and Rotated SAR Benchmark (CVPR 2025 · similarity 0.86)
- M3LEO: A Multi-Modal, Multi-Label Earth Observation Dataset Integrating Interferometric SAR and Multispectral Data (NeurIPS 2024 · similarity 0.85)
- FUSAR-GPT: A Spatiotemporal Feature-Embedded and Two-Stage Decoupled Visual Language Model for SAR Imagery (CVPR 2026 · similarity 0.83)
- Learning De-Biased Representations for Remote-Sensing Imagery (NeurIPS 2024 · similarity 0.83)
- Dual Information Purification for Lightweight SAR Object Detection (AAAI 2025 · similarity 0.83)
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.
