Research map / Detection and segmentation / Open-set and few-shot detection
No More Discrimination: Cross City Adaptation of Road Scene Segmenters
Yi-Hsin Chen, Wei-Yu Chen, Yu-Ting Chen, Bo-Cheng Tsai, Yu-Chiang Frank Wang, Min Sun
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 global domain-adversarial alignment (ablation) | Pre-trained dilated FCN · Cityscapes -> four cities (average) | mIoU: +2.6% mIoU | Improved |
| add class-wise alignment on top of global alignment (ablation) | GA · Cityscapes -> four cities (average) | mIoU: +0.9% mIoU | Improved |
| add static-object priors on top of GA+CA (ablation) | GA+CA · Cityscapes -> four cities (average) | mIoU: +0.6% mIoU | Improved |
| add static-object priors (ablation) | GA+CA · Cityscapes -> Tokyo | mIoU: 42.6 -> 42.8 | Improved |
| add static-object priors (ablation) | GA+CA · Cityscapes -> Rome | mIoU: 42.1 -> 42.9 | Improved |
| add static-object priors (ablation) | GA+CA · Cityscapes -> Rio | mIoU: 41.4 -> 42.5 | Improved |
| add static-object priors (ablation) | GA+CA · Cityscapes -> Taipei | mIoU: 38.8 -> 39.6 | Improved |
| add global alignment (ablation) | Pre-trained · SYNTHIA -> Cityscapes | mIoU: 30.7 -> 33.8 | Improved |
| add class-wise alignment on top of global alignment (ablation) | GA · SYNTHIA -> Cityscapes | mIoU: 33.8 -> 35.7 | Improved |
| unsupervised adaptation (Full Method) vs fully supervised fine-tuning | Full Method · Cityscapes -> Rome | mIoU: 42.9 vs 43.8 | No clear effect |
| apply segmenter trained on Frankfurt to farther cities without adaptation | Pre-trained dilated FCN · Cityscapes to other cities | mIoU: 64.6 -> 35.1 | Hurt |
Closest papers by meaning
- In Defense of Pre-Trained ImageNet Architectures for Real-Time Semantic Segmentation of Road-Driving Images (CVPR 2019 · similarity 0.89)
- Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-Modal Distillation (ECCV 2022 · similarity 0.88)
- A Curriculum Domain Adaptation Approach to the Semantic Segmentation of Urban Scenes (TPAMI 2019 · similarity 0.88)
- Unsupervised Semantic Segmentation of Urban Scenes via Cross-Modal Distillation (IJCV 2025 · similarity 0.88)
- Learning Content-Enhanced Mask Transformer for Domain Generalized Urban-Scene Segmentation (AAAI 2024 · similarity 0.86)
- Weakly Supervised Semantic Segmentation for Driving Scenes (AAAI 2024 · similarity 0.86)
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.
