Research map / Detection and segmentation / CNN features
RepPoints v2: Verification Meets Regression for Object Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 corner verification to RepPoints (ablation) | RepPoints · COCO val | AP: 39.1 -> 40.5 | Helped |
| Add corner verification to RepPoints (ablation) | RepPoints · COCO val | AP90: 14.4 -> 18.4 | Helped |
| Add foreground verification after corner verification (ablation) | RepPoints · COCO val | AP: 40.5 -> 41.0 | Helped |
| Add auxiliary verification for multi-task learning (ablation) | RepPoints · COCO val | AP: 39.1 -> 39.5 | Helped |
| Add feature enhancement to multi-task verification (ablation) | RepPoints · COCO val | AP: 39.5 -> 40.2 | Helped |
| Add joint inference to feature enhancement and multi-task verification (ablation) | RepPoints · COCO val | AP: 40.2 -> 41.0 | Helped |
| Use explicit-corners representation instead of min-max, partial min-max, or momentum conversion (ablation) | RepPoints with verification · COCO val | AP: 0.2-0.3 mAP better | Helped |
| Add verification fusion (ablation) | RepPoints · COCO val, 2x multi-scale, ResNet-50 | AP: 41.8 -> 43.9 | Helped |
| Add verification fusion (ablation) | RepPoints · COCO val, 2x multi-scale, ResNet-101 | AP: 43.4 -> 45.5 | Helped |
| Add verification fusion (ablation) | RepPoints · COCO val, 2x multi-scale, ResNeXt-101 | AP: 45.5 -> 47.3 | Helped |
| Add verification fusion | FCOS · COCO test-dev | AP: 38.2 -> 39.5 | Helped |
| Add contour and foreground verification | Dense RepPoints · COCO test-dev instance segmentation | mask AP: 37.6 -> 38.6 | Helped |
Closest papers by meaning
- Multiclass Confidence and Localization Calibration for Object Detection (CVPR 2023 · similarity 0.87)
- Bridging Precision and Confidence: A Train-Time Loss for Calibrating Object Detection (CVPR 2023 · similarity 0.86)
- Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignment (CVPR 2024 · similarity 0.86)
- PointOBB-v2: Towards Simpler, Faster, and Stronger Single Point Supervised Oriented Object Detection (ICLR 2025 · similarity 0.85)
- DiffuBox: Refining 3D Object Detection with Point Diffusion (NeurIPS 2024 · similarity 0.85)
- Unsupervised Object Detection with Theoretical Guarantees (NeurIPS 2024 · similarity 0.85)
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.