Research map / Detection and segmentation / Open-set and few-shot detection
Robust Domain Adaptive Object Detection With Unified Multi-Granularity Alignment
Libo Zhang, Wenzhang Zhou, Heng Fan, Tiejian Luo, Haibin Ling
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 OSGF to baseline (ablation) | FCOS VGG-16 baseline · Cityscapes to FoggyCityscapes ablation | mAP: 22.0% to 24.7% | Improved |
| add OSGF to MGA (ablation) | MGA FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 45.6% to 47.9% | Improved |
| add MGD to OSGF baseline (ablation) | FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 24.7% to 45.3% | Improved |
| replace EMA with AEMA, without OSGF (ablation) | MGD FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 44.4% to 45.6% | Improved |
| replace EMA with AEMA, with OSGF (ablation) | MGA FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 46.5% to 47.9% | Improved |
| add AEMA on top of OSGF+MGD (ablation) | FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 45.3% to 47.9% | Improved |
| add instance-level discriminator to pixel-level (ablation) | MGD · Cityscapes to FoggyCityscapes ablation | mAP: 40.6% to 44.7% | Improved |
| add category-level discriminator Dcat (ablation) | MGD · Cityscapes to FoggyCityscapes ablation | mAP: 44.7% to 47.9% | Improved |
| use Dcat instead of Dcen/Dgrp/Dcls (ablation) | MGA FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 1.7%, 1.9% and 2.7% | Improved |
| gated fusion instead of average fusion in OSGF (ablation) | OSGF FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 45.1 -> 47.9 | Improved |
| gated fusion instead of conv fusion in OSGF (ablation) | OSGF FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 46.2 -> 47.9 | Improved |
| new shared-conv OSGF instead of conference OSGF (ablation) | MGA FCOS VGG-16 · Cityscapes to FoggyCityscapes ablation | mAP: 46.8 -> 47.9 | Improved |
Closest papers by meaning
- Multi-Granularity Alignment Domain Adaptation for Object Detection (CVPR 2022 · similarity 0.97)
- Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment (AAAI 2025 · similarity 0.92)
- Sampling-Free Universal Domain Adaptive Object Detection with Synergistic Optimization Alignment (IJCV 2026 · similarity 0.92)
- Adapting Object Detectors via Selective Cross-Domain Alignment (CVPR 2019 · similarity 0.92)
- Domain Adaptive Object Detection via Asymmetric Tri-way Faster-RCNN (ECCV 2020 · similarity 0.91)
- Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignment (CVPR 2024 · similarity 0.89)
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.
