Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning from Scale-Invariant Examples for Domain Adaptation in Semantic Segmentation
M.Naseer Subhani, 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 |
|---|---|---|---|
| add focal loss to the scale-invariant self-training loss (ablation) | LSE (VGG16-FCN8) · GTA5 to Cityscapes | mIoU: 36.4 -> 39.0 | Improved |
| add focal loss to the scale-invariant self-training loss (ablation) | LSE (VGG16-FCN8) · SYNTHIA to Cityscapes (16 classes) | mIoU: 38.2 -> 40.0 | Improved |
| add focal loss to the scale-invariant self-training loss (ablation) | LSE (VGG16-FCN8) · SYNTHIA to Cityscapes (13 classes) | mIoU*: 44.9 -> 47.0 | Improved |
| add focal loss to balance class selection across rounds (ablation) | LSE · GTA5 to Cityscapes | images selected per class | Improved |
| use scale-invariant self-generated examples with focal loss instead of PyCDA | PyCDA (VGG16-FCN8) · SYNTHIA to Cityscapes (16 classes) | mIoU: 35.9 -> 40.0 | Improved |
| use scale-invariant self-generated examples with focal loss instead of PyCDA | PyCDA (VGG16-FCN8) · GTA5 to Cityscapes | mIoU: 37.2 -> 39.0 | Improved |
| use scale-invariant self-generated examples with focal loss instead of PyCDA | PyCDA (ResNet101) · GTA5 to Cityscapes | mIoU: 47.4 -> 47.5 | No clear effect |
| use scale-invariant self-generated examples with focal loss instead of PyCDA | PyCDA (ResNet101) · SYNTHIA to Cityscapes (16 classes) | mIoU: 46.7 -> 42.6 | Hurt |
| use LSE + FL instead of PyCDA, gap to oracle | PyCDA · GTA5 to Cityscapes performance gap (VGG16-FCN8) | gap to oracle (%): -27.9 -> -21.3 | Improved |
Closest papers by meaning
- Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training (ECCV 2018 · similarity 0.92)
- Unsupervised Intra-Domain Adaptation for Semantic Segmentation Through Self-Supervision (CVPR 2020 · similarity 0.92)
- Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (CVPR 2021 · similarity 0.92)
- Domain Adaptive Semantic Segmentation With Self-Supervised Depth Estimation (ICCV 2021 · similarity 0.92)
- Balanced Learning for Domain Adaptive Semantic Segmentation (ICML 2025 · similarity 0.91)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 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.
