Research map / Label-efficient and robust learning / Domain adaptation and generalization
Category Dictionary Guided Unsupervised Domain Adaptation for Object Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 CDG-UDA instead of HTCN | Faster R-CNN with VGG16 · Cityscapes to Foggy Cityscapes | mAP: 39.8 -> 42.3 | Helped |
| use CDG-UDA instead of PD | Faster R-CNN with ResNet101 · VOC to Watercolor | mAP: 56.9 -> 59.7 | Helped |
| use CDG-UDA instead of MTOR | cross-domain object detection · Sim10k to Cityscapes | mAP: 46.6 -> 48.8 | Helped |
| use category dictionary-guided pseudo-box annotation instead of softmax annotation (ablation) | CDG self-training · VOC to Watercolor | mAP: 55.3 -> 58 | Helped |
| add CDG self-training to source-only training (ablation) | VOC to Watercolor · VOC to Watercolor | mAP: 43.9 -> 58 | Helped |
| add a second CDG self-training stage (ablation) | CDG self-training · VOC to Watercolor | mAP: 58 -> 59.3 | Helped |
| add residual-based weights to pseudo-box training losses (ablation) | CDG self-training, stage 1 · VOC to Watercolor | mAP: 58 -> 58.5 | Helped |
| add residual-based weights to pseudo-box training losses (ablation) | CDG self-training, stage 2 · VOC to Watercolor | mAP: 59.3 -> 59.7 | Helped |
| vary the residual-weight parameter s (ablation) | CDG weighted self-training · VOC to Watercolor | mAP: 58.0, 58.1, 58.1, 58.3, 58.5, 58.3, 58.2 | Mixed |
| vary the source-target loss balance µ (ablation) | CDG weighted self-training · VOC to Watercolor | mAP: 55.9, 57.0, 58.0, 55.9 | Mixed |
| vary the pseudo-label threshold θ (ablation) | CDG-UDA · Sim10k to Cityscapes | mAP: 48.5, 48.8, 46.3, 45.7 | Mixed |
Closest papers by meaning
- Instance Relation Graph Guided Source-Free Domain Adaptive Object Detection (CVPR 2023 · similarity 0.91)
- Unified Domain Generalization and Adaptation for Multi-View 3D Object Detection (NeurIPS 2024 · similarity 0.88)
- Pseudo Label Refinery for Unsupervised Domain Adaptation on Cross-dataset 3D Object Detection (CVPR 2024 · similarity 0.87)
- Object-Aware Domain Generalization for Object Detection (AAAI 2024 · similarity 0.87)
- Mean Teacher DETR with Masked Feature Alignment: A Robust Domain Adaptive Detection Transformer Framework (AAAI 2024 · similarity 0.87)
- Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment (AAAI 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.