Research map / Detection and segmentation / CNN features
A Simple Baseline with Placement Prior for Point-Supervised Oriented Object Detection
Runxiang Liu, Kaikai Xie, Qianxi Cao, Yuming Fang, Jiebin Yan, Junjie Chen
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 Voronoi Watershed RBox loss (ablation) | Point2RBox-v2 · DOTA-v1.5 | mAP@0.5: 39.45 -> 40.65 | Improved |
| add Voronoi Watershed RBox loss (ablation) | Point2RBox-v2 · DroneVehicle | mAP@0.5: 53.70 -> 54.47 | Improved |
| add angle prior loss on top of RBox loss (ablation) | Point2RBox-v2 · DOTA-v1.5 | mAP@0.5: 40.65 -> 42.36 | Improved |
| add size prior loss on top of RBox loss (ablation) | Point2RBox-v2 · DOTA-v1.5 | mAP@0.5: 40.65 -> 42.08 | Improved |
| add angle prior loss on top of RBox loss (ablation) | PlacePriorBase · DroneVehicle | mAP@0.5: 38.21 -> 39.53 | Improved |
| add size prior loss on top of RBox loss (ablation) | PlacePriorBase · DOTA-v1.5 | mAP@0.5: 30.17 -> 31.42 | Improved |
| add size loss to RBox and angle losses (ablation) | PlacePriorBase · DOTA-v1.5 | mAP@0.5: 31.69 -> 32.38 | Improved |
| add all three loss terms (ablation) | Point2RBox-v2 · DroneVehicle | mAP@0.5: 53.70 -> 56.37 | Improved |
| add all three loss terms (ablation) | Point2RBox-v3 (end-to-end) · CODrone | mAP@0.5: 28.42 -> 30.43 | Improved |
| add all three loss terms (ablation) | PointOBB-v3 (end-to-end) · CODrone | mAP@0.5: 18.27 -> 19.38 | Mixed |
| add point annotation noise sigma 30% vs 0% (ablation) | PlacePrior+P2R2 · DroneVehicle | mAP@0.5: 56.37 -> 53.95 | Hurt |
| add point annotation noise sigma 30% vs 0% (ablation) | PointOBB-v3 · DroneVehicle | mAP@0.5: 42.43 -> 36.38 | Hurt |
Closest papers by meaning
- PointOBB: Learning Oriented Object Detection via Single Point Supervision (CVPR 2024 · similarity 0.89)
- ABBSPO: Adaptive Bounding Box Scaling and Symmetric Prior based Orientation Prediction for Detecting Aerial Image Objects (CVPR 2025 · similarity 0.89)
- PointSR: Self-Regularized Point Supervision for Drone-View Object Detection (CVPR 2025 · similarity 0.88)
- FRED: Towards a Full Rotation-Equivariance in Aerial Image Object Detection (AAAI 2024 · similarity 0.88)
- PointOBB-v2: Towards Simpler, Faster, and Stronger Single Point Supervised Oriented Object Detection (ICLR 2025 · similarity 0.87)
- Point2RBox: Combine Knowledge from Synthetic Visual Patterns for End-to-end Oriented Object Detection with Single Point Supervision (CVPR 2024 · similarity 0.86)
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.
