Research map / Detection and segmentation / CNN features
Siamese DETR
Zeren Chen, Gengshi Huang, Wei Li, Jianing Teng, Kun Wang, Jing Shao, Chen Change Loy, Lu Sheng
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 |
|---|---|---|---|
| pretrain with Siamese multi-view region detection and semantic discrimination | DETR variants · COCO val2017 | AP: 40.2 -> 40.5; 41.5 -> 43.0; 40.8 -> 42.1; 45.5 -> 46.3 | Improved |
| pretrain with Siamese multi-view region detection and semantic discrimination | DETR variants · PASCAL VOC test2007 | AP: 57.5 -> 58.1; 55.9 -> 57.7; 54.0 -> 58.0; 59.7 -> 61.2 | Improved |
| extend single-view patch detection to multi-view region detection (ablation) | Conditional DETR · PASCAL VOC test2007 | AP: 56.9 -> 57.1 | Improved |
| add global semantic discrimination to multi-view region detection (ablation) | Conditional DETR · PASCAL VOC test2007 | AP: 57.1 -> 57.3 | Improved |
| add object-level region discrimination to multi-view region detection (ablation) | Conditional DETR · PASCAL VOC test2007 | AP: 57.1 -> 57.3 | Improved |
| use crop-level instead of object-level region features for region discrimination (ablation) | Conditional DETR · PASCAL VOC test2007 | AP: 57.3 -> 57.7 | Improved |
| add global discrimination to crop-level region discrimination (ablation) | Conditional DETR · PASCAL VOC test2007 | AP: 57.7 -> 58.1 | Improved |
| replace EdgeBoxes with random proposals (ablation) | Siamese DETR · COCO val2017 (ImageNet pretraining) | AP: 40.5 -> 40.4 | No clear effect |
| replace EdgeBoxes with random proposals (ablation) | Siamese DETR · PASCAL VOC test2007 (COCO pretraining) | AP: 57.7 -> 54.9 | Hurt |
| replace EdgeBoxes with Selective Search proposals (ablation) | Siamese DETR · PASCAL VOC test2007 (COCO pretraining) | AP: 57.7 -> 56.2 | Hurt |
Closest papers by meaning
- Siamese Image Modeling for Self-Supervised Vision Representation Learning (CVPR 2023 · similarity 0.87)
- DisCo DETR: Distance-aware Multi-view Contrastive Learning for DETR Pre-training (AAAI 2026 · similarity 0.87)
- Exploring the Equivalence of Siamese Self-Supervised Learning via a Unified Gradient Framework (CVPR 2022 · similarity 0.87)
- FP-DETR: Detection Transformer Advanced by Fully Pre-training (ICLR 2022 · similarity 0.86)
- Distractor-aware Siamese Networks for Visual Object Tracking (ECCV 2018 · similarity 0.86)
- Mr. DETR: Instructive Multi-Route Training for Detection Transformers (CVPR 2025 · 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.
