Research map / Detection and segmentation / Weakly supervised detection and segmentation
DTG-SSOD: Dense Teacher Guidance for Semi-Supervised 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 |
|---|---|---|---|
| apply dense teacher guidance to RPN (ablation) | supervised baseline · COCO 10% | mAP: 26.8 -> 28.5 | Helped |
| also apply dense teacher guidance to RCNN (ablation) | DTG-SSOD · COCO 10% | mAP: 28.5 -> 36.3 | Helped |
| add INC classification labels (ablation) | DTG-SSOD · COCO 10% | mAP: 28.5 -> 32.9 | Helped |
| add INC regression labels (ablation) | DTG-SSOD · COCO 10% | mAP: 32.9 -> 35.1 | Helped |
| add Rank Matching (ablation) | DTG-SSOD · COCO 10% | mAP: 35.1 -> 36.3 | Helped |
| dense-to-dense instead of sparse-to-dense in RCNN (ablation) | DTG-SSOD · COCO 10% | mAP: 34.5 -> 35.5 | Helped |
| dense-to-dense instead of sparse-to-dense in RPN (ablation) | DTG-SSOD · COCO 10% | mAP: 35.5 -> 36.3 | Helped |
| dense-to-dense instead of sparse-to-dense at 1% labels (ablation) | DTG-SSOD · COCO 1% | mAP: 21.1 -> 21.3 | No clear effect |
| dense-to-dense instead of sparse-to-dense at 5% and 10% labels (ablation) | DTG-SSOD · COCO 5%/10% | mAP: 31.1 -> 31.9 (5%); 34.5 -> 36.3 (10%) | Helped |
| regression target: reserved box instead of all teacher boxes (ablation) | INC · COCO 10% | mAP: 34.3 -> 35.1 | Helped |
| vary temperature T (0.5, 1.0, 5.0) (ablation) | Rank Matching · COCO 10% | mAP: 35.6 / 35.8 / 35.6 | No clear effect |
| vary RM loss weight beta (1, 3, 5) (ablation) | DTG-SSOD · COCO 10% | mAP: 35.8 / 36.3 / 36.1 | Mixed |
Closest papers by meaning
- Dense Learning Based Semi-Supervised Object Detection (CVPR 2022 · similarity 0.92)
- Humble Teachers Teach Better Students for Semi-Supervised Object Detection (CVPR 2021 · similarity 0.92)
- Active Teacher for Semi-Supervised Object Detection (CVPR 2022 · similarity 0.91)
- Unbiased Teacher for Semi-Supervised Object Detection (ICLR 2021 · similarity 0.90)
- Sparse Semi-DETR: Sparse Learnable Queries for Semi-Supervised Object Detection (CVPR 2024 · similarity 0.88)
- A-Teacher: Asymmetric Network for 3D Semi-Supervised Object Detection (CVPR 2024 · similarity 0.87)
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.