Research map / Detection and segmentation / CNN features
DBF: Dynamic Belief Fusion for Combining Multiple Object Detectors
Hyungtae Lee, Heesung Kwon
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 |
|---|---|---|---|
| use dynamic basic probability assignment instead of static DST (ablation) | DBF · PASCAL VOC 07 (six R-CNNs) | mAP: .601 -> .760 | Improved |
| add an intermediate hypothesis (as in DBF) over naive Bayesian fusion (ablation) | DBF · PASCAL VOC 07 (8-detector pool) | mAP: .354 -> .553 | Improved |
| add an intermediate hypothesis (as in DBF) over naive Bayesian fusion (ablation) | DBF · PASCAL VOC 07 (pool without R-CNN) | mAP: .253 -> .341 | Improved |
| fuse detectors with DBF instead of best individual R-CNN | DBF · PASCAL VOC 07 (six R-CNNs) | mAP: .739 -> .760 | Improved |
| fuse detectors with DBF instead of best individual R-CNN | DBF · PASCAL VOC 12 (six R-CNNs) | mAP: .727 -> .739 | Improved |
| use ideal theoretical detector (n = infinity) instead of tuned shape n (ablation) | DBF · PASCAL VOC 07 (six R-CNNs) | mAP | Hurt |
| increase number of fused detectors from 2 to 4 (ablation) | DBF · ARL | mAP: .295 -> .325 | Improved |
| increase number of fused detectors from 2 to 8 (ablation) | DBF · PASCAL VOC 07 (8-detector pool) | mAP: .545 -> .553 | Improved |
| increase number of fused detectors from 2 to 6 (ablation) | DBF · PASCAL VOC 07 (six R-CNNs) | mAP: .742 -> .760 | Improved |
| use trainval/trainval partition instead of train/val (ablation) | DBF · PASCAL VOC 07 (six R-CNNs) | mAP: .709 -> .760 | Improved |
Closest papers by meaning
- CRT-Fusion: Camera, Radar, Temporal Fusion Using Motion Information for 3D Object Detection (NeurIPS 2024 · similarity 0.84)
- Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion (ICLR 2025 · similarity 0.84)
- Neural Exposure Fusion for High-Dynamic Range Object Detection (CVPR 2024 · similarity 0.83)
- Test-Time Dynamic Image Fusion (NeurIPS 2024 · similarity 0.82)
- Predictive Dynamic Fusion (ICML 2024 · similarity 0.82)
- Test-Time Adaptive Object Detection with Foundation Model (NeurIPS 2025 · similarity 0.82)
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.
