Research map / Label-efficient and robust learning / Domain adaptation and generalization
ML-BPM: Multi-Teacher Learning with Bidirectional Photometric Mixing for Open Compound Domain Adaptation in Semantic Segmentation
Fei Pan, Sungsu Hur, Seokju Lee, Junsik Kim, In So Kweon
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 multi-teacher learning (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 38.0 -> 40.2 | Improved |
| add mixing in the alpha direction (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 38.5 -> 40.2 | Improved |
| add mixing in the beta direction (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 38.9 -> 40.2 | Improved |
| add subdomain style purification (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 39.7 -> 40.2 | Improved |
| add adaptive distillation (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 39.6 -> 40.2 | Improved |
| add multi-teacher learning to DACS (ablation) | DACS · GTA5->C-Driving (val) | mIoU: 36.6 -> 39.1 | Improved |
| add bidirectional mixing to DACS (ablation) | DACS · GTA5->C-Driving (val) | mIoU: 36.6 -> 37.3 | Improved |
| add photometric mixing to DACS (ablation) | DACS · GTA5->C-Driving (val) | mIoU: 36.6 -> 37.4 | Improved |
| add bidirectional photometric mixing to DACS (ablation) | DACS · GTA5->C-Driving (val) | mIoU: 36.6 -> 37.8 | Improved |
| use BPM instead of ClassMix for mixing (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 39.1 -> 40.2 | Improved |
| use BPM instead of CutMix (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 37.6 -> 40.2 | Improved |
| use BPM instead of CowMix (ablation) | ML-BPM · GTA5->C-Driving (val) | mIoU: 37.4 -> 40.2 | Improved |
Closest papers by meaning
- Cluster, Split, Fuse, and Update: Meta-Learning for Open Compound Domain Adaptive Semantic Segmentation (CVPR 2021 · similarity 0.91)
- Discover, Hallucinate, and Adapt: Open Compound Domain Adaptation for Semantic Segmentation (NeurIPS 2020 · similarity 0.90)
- Bidirectional Domain Mixup for Domain Adaptive Semantic Segmentation (AAAI 2023 · similarity 0.90)
- Learning to Adapt via Latent Domains for Adaptive Semantic Segmentation (NeurIPS 2021 · similarity 0.90)
- Multi-Target Domain Adaptation With Collaborative Consistency Learning (CVPR 2021 · similarity 0.90)
- Open-Set Domain Adaptation for Semantic Segmentation (CVPR 2024 · similarity 0.89)
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.
