Research map / Detection and segmentation / CNN features
SAP-DETR: Bridging the Gap Between Salient Points and Queries-Based Transformer Detector for Fast Model Convergency
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 |
|---|---|---|---|
| Replace center-based DAB-DETR priors with query-specific salient points (ablation) | DAB-DETR · COCO val2017, R50, 12 epochs | AP: 32.3 to 33.5 | Helped |
| Add movable reference points (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP: 35.2 to 36.2 | Helped |
| Add movable reference points (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP_S: 15.8 to 16.4 | Helped |
| Add inner matching loss (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP_75: 37.4 to 37.9 | Helped |
| Add inner matching loss (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP_50: 56.3 to 56.2 | Hurt |
| Add PECA (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP: 34.8 to 36.2 | Helped |
| Add SDG (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP: 35.6 to 36.2 | Helped |
| Add SDG to a model already using PECA (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP_S: 16.3 to 16.4 | No clear effect |
| Combine inner matching loss with movable-within-grid updates (ablation) | SAP-DETR · COCO val2017, R50, 12 epochs | AP: 26.3 to 36.2 | Helped |
| Add SAP design to DN-DETR | DN-DETR · COCO val2017, R50, 12 epochs | AP: 38.3 -> 39.5 | Helped |
| Add SAP design to DINO | DINO · COCO val2017, R50, 12 epochs | AP: 39.7 -> 40.0 | Helped |
| Add SAP design to Group-DETR | Group-DETR · COCO val2017, R50, 12 epochs | AP: 39.1 -> 39.8 | Helped |
Closest papers by meaning
- Salience DETR: Enhancing Detection Transformer with Hierarchical Salience Filtering Refinement (CVPR 2024 · similarity 0.89)
- KD-DETR: Knowledge Distillation for Detection Transformer with Consistent Distillation Points Sampling (CVPR 2024 · similarity 0.89)
- State Space Model Meets Transformer: A New Paradigm for 3D Object Detection (ICLR 2025 · similarity 0.87)
- MI-DETR: An Object Detection Model with Multi-time Inquiries Mechanism (CVPR 2025 · similarity 0.87)
- DETRs Beat YOLOs on Real-time Object Detection (CVPR 2024 · similarity 0.87)
- Hybrid Proposal Refiner: Revisiting DETR Series from the Faster R-CNN Perspective (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.