Research map / Detection and segmentation / Open-set and few-shot detection
DLDA: Unified Dual-Level Domain Adaptation for Low-Light Object Detection
Jiayi Hu, Qian Zhao, Gang 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 |
|---|---|---|---|
| use DLDA instead of source-only YOLOv5-L | YOLOv5-L · COCO->ExDark (unsupervised DA, test) | mAP@50: 63.4 -> 68.8 | Improved |
| use DLDA instead of DAI | DAI · COCO->ExDark (unsupervised DA, test) | mAP@50: 66.7 -> 68.8 | Improved |
| use DLDA instead of DAI for face detection | DAI · WIDER FACE->DARK FACE (unsupervised DA) | mAP: 26.8 -> 27.5 | Improved |
| add DISP low-light synthesis alone (ablation) | YOLOv5-L source only · COCO->ExDark ablation (mAP@50) | mAP@50: 63.4 -> 62.2 | Hurt |
| add CUT translation alone (ablation) | YOLOv5-L source only · COCO->ExDark ablation (mAP@50) | mAP@50: 63.4 -> 65.5 | Improved |
| add LCT translation alone (ablation) | YOLOv5-L source only · COCO->ExDark ablation (mAP@50) | mAP@50: 63.4 -> 66.3 | Improved |
| add MSCA alone (ablation) | YOLOv5-L source only · COCO->ExDark ablation (mAP@50) | mAP@50: +2.5 | Improved |
| combine LCT and MSCA (ablation) | DLDA · COCO->ExDark ablation (mAP@50) | mAP@50: 66.3 -> 68.8 | Improved |
Closest papers by meaning
- Self-Guided Low Light Object Detection Framework (ICLR 2026 · similarity 0.91)
- Unsupervised Face Detection in the Dark (TPAMI 2022 · similarity 0.89)
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation (CVPR 2024 · similarity 0.88)
- Adapting Object Detectors via Selective Cross-Domain Alignment (CVPR 2019 · similarity 0.88)
- You Only Look Around: Learning Illumination-Invariant Feature for Low-light Object Detection (NeurIPS 2024 · similarity 0.87)
- FRBNet: Revisiting Low-Light Vision through Frequency-Domain Radial Basis Network (NeurIPS 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.
