Research map / Detection and segmentation / Open-set and few-shot detection
SSAL: Synergizing between Self-Training and Adversarial Learning for Domain Adaptive Object Detection
Muhammad Akhtar Munir, Muhammad Haris Khan, M. Saquib Sarfraz, Mohsen Ali
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 |
|---|---|---|---|
| combine uncertainty-guided pseudo-labeling and uncertainty-guided tiling | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 46.0 -> 51.8 | Improved |
| add uncertainty-guided pseudo-labeling (UGPL) (ablation) | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 46.0 -> 49.5 | Improved |
| add uncertainty-guided tiling (UGT) (ablation) | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 46.0 -> 50.0 | Improved |
| use confidence-based pseudo-label selection instead of uncertainty-guided selection (ablation) | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 49.5 -> 43.2 | Hurt |
| use uncertainty-guided tiles instead of full images, random tiles, or certain tiles (ablation) | UGPL adversarial alignment · Sim10K → Cityscapes | AP@0.5: 48.1 / 49.8 / 50.2 -> 51.8 | Improved |
| add R1 and R2 rounds with both UGPL and UGT after R0 (ablation) | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 46.3 -> 51.8 | Improved |
| increase spatial dropout from 10% to 20% or 30% (ablation) | SSAL · Sim10K → Cityscapes | AP@0.5: 51.8 -> 50.3; 51.8 -> 49.4 | Hurt |
| vary uncertainty threshold κ₁ from 0.5 to 0.4 or 0.6 (ablation) | SSAL · Sim10K → Cityscapes | AP (mean): 28.9 -> 28.6; 28.9 -> 28.3 | Hurt |
| increase IoU threshold γ from 0.5 to 0.6 or 0.7 (ablation) | SSAL · Sim10K → Cityscapes | AP (mean): 28.9 -> 28.3; 28.9 -> 27.5 | Hurt |
| adapt with SSAL rather than EPM | FCOS domain adaptation · Sim10K → Cityscapes | AP@0.5: 49.0 -> 51.8 | Improved |
| adapt with SSAL rather than EPM | FCOS domain adaptation · KITTI → Cityscapes | AP@0.5: 43.2 -> 45.6 | Improved |
| adapt with SSAL rather than the strongest reported baseline | FCOS domain adaptation · Cityscapes → Foggy Cityscapes | mAP@0.5: 40.9 -> 39.6 | Hurt |
Closest papers by meaning
- Domain Adaptive Object Detection via Balancing Between Self-Training and Adversarial Learning (TPAMI 2023 · similarity 0.97)
- Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment (AAAI 2025 · similarity 0.89)
- Toward Improving Robustness and Accuracy in Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.89)
- Mean Teacher DETR with Masked Feature Alignment: A Robust Domain Adaptive Detection Transformer Framework (AAAI 2024 · similarity 0.88)
- Large Self-Supervised Models Bridge the Gap in Domain Adaptive Object Detection (CVPR 2025 · similarity 0.88)
- Differential Alignment for Domain Adaptive Object Detection (AAAI 2025 · similarity 0.88)
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.
