Research map / Detection and segmentation / Open-set and few-shot detection
Reasoning-RCNN: Unifying Adaptive Global Reasoning Into Large-Scale Object Detection
Check 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 relation-knowledge adaptive global reasoning | Faster R-CNN with FPN · VG1000 test | AP: 7.1 -> 8.2 | Helped |
| Add relation-knowledge adaptive global reasoning | Faster R-CNN with FPN · VG3000 test | AP: 3.7 -> 4.3 | Helped |
| Add relation-knowledge adaptive global reasoning | Faster R-CNN with FPN · ADE test | AP: 11.3 -> 15.5 | Helped |
| Add relation-knowledge adaptive global reasoning | Faster R-CNN · PASCAL VOC 2007 test | mAP: 80.8 -> 82.5 | Helped |
| Add relation-knowledge adaptive global reasoning | Faster R-CNN with FPN · MS COCO 2017 | mAP: 37.3 -> 42.9 | Helped |
| Replace the commonsense knowledge graph with an identity matrix (ablation) | Reasoning-RCNN with FPN · MS COCO 2017 | mAP: 42.9 -> 39.5 (−3.4) | Hurt |
| Use the global semantic pool instead of region features (ablation) | Reasoning-RCNN with FPN · MS COCO 2017 | mAP: 40.2 -> 42.9 (−2.7 for region-feature variant) | Helped |
| Add image-wise adaptive attention (ablation) | Reasoning-RCNN with FPN · MS COCO 2017 | mAP: 40.9 -> 42.9 (−2.0 without attention) | Helped |
| Use image-wise adaptive attention instead of uniform category weights (ablation) | Reasoning-RCNN with FPN · MS COCO 2017 | mAP: 41.5 -> 42.9 (−1.4 for uniform weights) | Helped |
| Add attribute-knowledge adaptive global reasoning | Faster R-CNN · MS COCO 2017 | mAP: 34.9 -> 39.2 | Helped |
Closest papers by meaning
- Hybrid Knowledge Routed Modules for Large-scale Object Detection (NeurIPS 2018 · similarity 0.91)
- Universal-RCNN: Universal Object Detector via Transferable Graph R-CNN (AAAI 2020 · similarity 0.90)
- Spatially Invariant Unsupervised Object Detection with Convolutional Neural Networks (AAAI 2019 · similarity 0.89)
- R-FCN: Object Detection via Region-based Fully Convolutional Networks (NeurIPS 2016 · similarity 0.88)
- Large Scale Semi-Supervised Object Detection Using Visual and Semantic Knowledge Transfer (CVPR 2016 · similarity 0.88)
- GCD: Advancing Vision-Language Models for Incremental Object Detection via Global Alignment and Correspondence Distillation (AAAI 2025 · 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.