Research map / Detection and segmentation / Open-set and few-shot detection
SFUOD: Source-Free Unknown Object Detection
Keon-Hee Park, Seun-An Choe, Gyeong-Moon Park
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 |
|---|---|---|---|
| add collaborative tuning to DRU (ablation) | DRU · Cityscapes -> Foggy Cityscapes | Known mAP: 22.97 -> 30.63 | Improved |
| add PAUL to DRU (ablation) | DRU · Cityscapes -> Foggy Cityscapes | U-Recall: 3.60 -> 6.46 | Improved |
| add PAUL on top of collaborative tuning (ablation) | CollaPAUL · Cityscapes -> Foggy Cityscapes | H-Score: 6.38 -> 15.95 | Improved |
| add objectness mask to confidence mask in PAUL (ablation) | PAUL · Cityscapes -> Foggy Cityscapes | U-Recall: 3.56 -> 10.59 | Improved |
| add confidence mask to objectness mask in PAUL (ablation) | PAUL · Cityscapes -> Foggy Cityscapes | H-Score: 14.59 -> 15.95 | Improved |
| use cross-domain attention instead of prefix-tuning (ablation) | collaborative tuning · Cityscapes -> Foggy Cityscapes | H-Score: 12.57 -> 15.95 | Improved |
| use PAUL instead of attention-driven unknown labeling (ablation) | unknown labeling · Cityscapes -> Foggy Cityscapes | U-Recall: 4.00 -> 10.59 | Improved |
| use 3 collaborative layers instead of 1 (ablation) | CollaPAUL · Cityscapes -> Foggy Cityscapes | H-Score: 10.96 -> 15.95 | Improved |
| use 5 collaborative layers instead of 3 (ablation) | CollaPAUL · Cityscapes -> Foggy Cityscapes | H-Score: 15.95 -> 13.02 | Hurt |
| use 4 collaborative layers instead of 3 (ablation) | CollaPAUL · Cityscapes -> Foggy Cityscapes | Known mAP: 32.32 -> 32.90 | No clear effect |
| use CollaPAUL instead of DRU | DRU · Cityscapes -> Foggy Cityscapes | H-Score: 6.22 -> 15.95 | Improved |
| use CollaPAUL instead of DRU | DRU · Cityscapes -> BDD100K | H-Score: 9.82 -> 13.15 | Improved |
Closest papers by meaning
- Simplifying Source-Free Domain Adaptation for Object Detection: Effective Self-Training Strategies and Performance Insights (ECCV 2024 · similarity 0.90)
- Beyond Boundaries: Leveraging Vision Foundation Models for Source-Free Object Detection (AAAI 2026 · similarity 0.89)
- Enhancing Source-Free Domain Adaptive Object Detection with Low-confidence Pseudo Label Distillation (ECCV 2024 · similarity 0.88)
- Instance Relation Graph Guided Source-Free Domain Adaptive Object Detection (CVPR 2023 · similarity 0.88)
- Source-Free Object Detection by Learning To Overlook Domain Style (CVPR 2022 · similarity 0.88)
- Dynamic Retraining-Updating Mean Teacher for Source-Free Object Detection (ECCV 2024 · similarity 0.88)
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.
