Research map / Detection and segmentation / Open-set and few-shot detection
Detecting Unknown Objects via Energy-based Separation for Open World Object Detection
Jun-Woo Heo, Keonhee Park, 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 EUS (ETF-subspace unknown separation) (ablation) | OrthogonalDet baseline with improved pseudo-labels · M-OWODB Task 1 | U-Recall: 36.8 -> 65.1 | Improved |
| add EUS (ablation) | baseline · M-OWODB Task 3 | U-Recall: 30.3 -> 69.0 | Improved |
| add EUS (ablation) | baseline · M-OWODB Task 4 | Known mAP: 44.7 -> 43.5 | Mixed |
| add EUS (ablation) | baseline · M-OWODB Task 1 | H-Score: 47.2 -> 65.6 | Improved |
| add EKD loss (ablation) | baseline · M-OWODB Task 4 | Known mAP: 44.7 -> 45.9 | Improved |
| add EKD loss (ablation) | baseline · M-OWODB Task 3 | U-Recall: 30.3 -> 38.9 | Improved |
| add EKD loss on top of EUS (ablation) | EUS · M-OWODB Task 4 | Known mAP: 43.5 -> 46.0 | Improved |
| add EKD loss on top of EUS (ablation) | EUS · M-OWODB Task 3 | H-Score: 57.5 -> 58.0 | Improved |
| add EKD loss on top of EUS (ablation) | EUS · M-OWODB Task 2 | Known mAP: 51.9 -> 53.3 | Improved |
| add EUS on top of EKD (ablation) | EKD · M-OWODB Task 3 | H-Score: 43.9 -> 58.0 | Improved |
| EKD loss changes task-head energy scores | DEUS · M-OWODB | energy score T3-T3: 0.89 -> 1.01 | Improved |
| EKD loss suppresses off-diagonal energy | DEUS · M-OWODB | energy score T1-T3: -3.63 -> -7.20 | Improved |
Closest papers by meaning
- Looking Beyond the Known: Towards a Data Discovery Guided Open-World Object Detection (NeurIPS 2025 · similarity 0.90)
- Open-World Objectness Modeling Unifies Novel Object Detection (CVPR 2025 · similarity 0.89)
- OW-OVD: Unified Open World and Open Vocabulary Object Detection (CVPR 2025 · similarity 0.88)
- Hyp-OW: Exploiting Hierarchical Structure Learning with Hyperbolic Distance Enhances Open World Object Detection (AAAI 2024 · similarity 0.88)
- Percept, Memory, and Imagine: World Feature Simulating for Open-Domain Unknown Object Detection (CVPR 2025 · similarity 0.87)
- Detecting Open World Objects via Partial Attribute Assignment (CVPR 2025 · 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.
