Research map / Detection and segmentation / Open-set and few-shot detection
An Empirical Study on Multi-domain Robust Semantic Segmentation
Yajie Liu, Pu Ge, Qingjie Liu, Shichao Fan, Yunhong Wang
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 DomainMix (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.07 -> 64.51 | Improved |
| add CutMix (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.07 -> 64.54 | Improved |
| add Copy-Paste (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.07 -> 64.19 | No clear effect |
| combine Copy-Paste and DomainMix (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.07 -> 64.82 | Improved |
| add CutMix on top of Copy-Paste and DomainMix (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.82 -> 64.50 | Hurt |
| add class-balanced sampling (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.1 -> 64.2 | No clear effect |
| add warmup (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.1 -> 64.2 | No clear effect |
| combine class-balanced sampling and warmup (ablation) | Segformer-MiT-B5 · RVC six datasets | h.mean mIoU: 64.1 -> 64.8 | Improved |
| add domain generalization method IRM (ablation) | DeepLabv3+/ResNet50 · RVC six datasets | h.mean mIoU: 49.6 -> 37.7 | Hurt |
| add domain generalization method DANN (ablation) | DeepLabv3+/ResNet50 · RVC six datasets | h.mean mIoU: 49.6 -> 48.1 | Hurt |
| train on six datasets instead of CityScapes alone (ablation) | Segformer-MiT-B1 · CityScapes | mIoU: 72.0 -> 70.4 | Hurt |
| train on six datasets instead of CityScapes alone (ablation) | Segformer-MiT-B5 · CityScapes | mIoU: 76.9 -> 76.4 | No clear effect |
Closest papers by meaning
- Scaling Up Multi-domain Semantic Segmentation with Sentence Embeddings (IJCV 2024 · similarity 0.89)
- Pulling Target to Source: A New Perspective on Domain Adaptive Semantic Segmentation (IJCV 2024 · similarity 0.89)
- Unsupervised Intra-Domain Adaptation for Semantic Segmentation Through Self-Supervision (CVPR 2020 · similarity 0.88)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · similarity 0.88)
- SemiDAViL: Semi-supervised Domain Adaptation with Vision-Language Guidance for Semantic Segmentation (CVPR 2025 · similarity 0.88)
- Exploring Semantic Consistency and Style Diversity for Domain Generalized Semantic Segmentation (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.
