Research map / Label-efficient and robust learning / Domain adaptation and generalization
Source-Free Domain Adaptation for Semantic Segmentation
Check 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 source-free knowledge transfer and target adaptation with IPSM | DeepLabV3 · GTA5 → Cityscapes | mIoU: 43.12 -> 43.16 | Helped |
| Add IPSM (ablation) | SFDA with DeepLabV3 · GTA5 → Cityscapes | mIoU: 41.35 -> 43.16 | Helped |
| Add IPSM (ablation) | SFDA with SegNet · GTA5 → Cityscapes | mIoU: 34.43 -> 35.86 | Helped |
| Add IPSM (ablation) | SFDA · SYNTHIA → Cityscapes | mIoU: 37.50 -> 39.20 | Helped |
| Combine BNS and DAD losses instead of using BNS alone (ablation) | SFKT · GTA5 component study | score: 49.8 -> 58.3 | Helped |
| Use DAD loss rather than BNS loss alone (ablation) | SFKT · GTA5 component study | score: 49.8 -> 55.4 | Helped |
| Combine BNS and DAD losses instead of using BNS alone (ablation) | SFKT · Cityscapes component study | score: 60.6 -> 70.8 | Helped |
| Use DAD loss rather than BNS loss alone (ablation) | SFKT · Cityscapes component study | score: 60.6 -> 65.5 | Helped |
| Combine BNS and DAD losses instead of using BNS alone (ablation) | SFKT · SYNTHIA component study | score: 51.4 -> 59.0 | Helped |
| Use DAD loss rather than BNS loss alone (ablation) | SFKT · SYNTHIA component study | score: 51.4 -> 54.7 | Helped |
| Set α below or above 1.0 (ablation) | SFDA · GTA5 → Cityscapes | mIoU: 41.33, 42.70, 43.16, 42.47 | Mixed |
| Set β below or above 0.5 (ablation) | SFDA · GTA5 → Cityscapes | mIoU: 41.54, 43.09, 43.16, 42.47 | Mixed |
Closest papers by meaning
- Unsupervised Model Adaptation for Continual Semantic Segmentation (AAAI 2021 · similarity 0.93)
- Spatio-Temporal Pixel-Level Contrastive Learning-Based Source-Free Domain Adaptation for Video Semantic Segmentation (CVPR 2023 · similarity 0.93)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · similarity 0.93)
- Open-Set Domain Adaptation for Semantic Segmentation (CVPR 2024 · similarity 0.92)
- Balanced Learning for Domain Adaptive Semantic Segmentation (ICML 2025 · similarity 0.91)
- Towards Unsupervised Domain Bridging via Image Degradation in Semantic Segmentation (NeurIPS 2025 · similarity 0.90)
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.