Research map / Detection and segmentation / Open-set and few-shot detection
DA-Mamba: Learning Domain-Aware State Space Model for Global-Local Alignment in Domain Adaptive Object Detection
Haochen Li, Rui Zhang, Hantao Yao, Xin Zhang, Yifan Hao, Shaohui Peng, Yongwei Zhao, Ling Li
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 IA-SSM to the backbone (ablation) | YOLO-World UDA baseline · Cityscapes->Foggy Cityscapes | mAP@0.5: 52.3 -> 56.8 | Improved |
| add OA-SSM to the detection head (ablation) | YOLO-World UDA baseline · Cityscapes->Foggy Cityscapes | mAP@0.5: 52.3 -> 55.4 | Improved |
| add both IA-SSM and OA-SSM (ablation) | YOLO-World UDA baseline · Cityscapes->BDD100K | mAP@0.5: 41.9 -> 48.7 | Improved |
| use IA-SSM instead of standard Mamba in backbone (ablation) | DA-Mamba · Cityscapes->Foggy Cityscapes | mAP@0.5: 54.3 -> 56.8 | Improved |
| use OA-SSM instead of standard Mamba in head (ablation) | DA-Mamba · Cityscapes->Foggy Cityscapes | mAP@0.5: 53.4 -> 55.4 | Improved |
| use DA-Mamba modules instead of standard Mamba in both (ablation) | DA-Mamba · Cityscapes->Foggy Cityscapes | mAP@0.5: 55.0 -> 58.1 | Improved |
| add convolution pipeline to the SSM pipeline of IA-SSM (ablation) | IA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 54.3 -> 55.6 | Improved |
| add SSM pipeline to the convolution pipeline of IA-SSM (ablation) | IA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 53.5 -> 55.6 | Improved |
| add learnable visual prompts to IA-SSM (ablation) | IA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 55.6 -> 56.8 | Improved |
| add learnable visual prompts to OA-SSM (ablation) | OA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 54.8 -> 55.4 | Improved |
| attach IA-SSM at C5 instead of C3 (ablation) | IA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 53.8 -> 54.3 | Improved |
| use parallel instead of serial SSM structure (ablation) | IA-SSM/OA-SSM · Cityscapes->Foggy Cityscapes | mAP@0.5: 58.1 -> 58.1 | No clear effect |
Closest papers by meaning
- Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment (AAAI 2025 · similarity 0.92)
- DA-Ada: Learning Domain-Aware Adapter for Domain Adaptive Object Detection (NeurIPS 2024 · similarity 0.90)
- SEEN-DA: SEmantic ENtropy guided Domain-aware Attention for Domain Adaptive Object Detection (CVPR 2025 · similarity 0.89)
- Towards Unsupervised Model Selection for Domain Adaptive Object Detection (NeurIPS 2024 · similarity 0.89)
- Large Self-Supervised Models Bridge the Gap in Domain Adaptive Object Detection (CVPR 2025 · similarity 0.89)
- DSD-DA: Distillation-based Source Debiasing for Domain Adaptive Object Detection (ICML 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.
