Research map / Point clouds / Segmentation and detection
Geometry-Aware Network for Domain Adaptive Semantic Segmentation
Yinghong Liao, Wending Zhou, Xu Yan, Zhen Li, Yizhou Yu, Shuguang Cui
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 GANDA geometry-aware adaptation | HRDA · GTA5 → Cityscapes | mIoU: 73.8 -> 74.5 | Improved |
| Add GANDA geometry-aware adaptation | HRDA · GTA5 → Cityscapes | s-mIoU: 68.2 -> 69.0 | Improved |
| Add GANDA geometry-aware adaptation | HRDA · SYNTHIA → Cityscapes | mIoU: 65.8 -> 67.0 | Improved |
| Add GANDA geometry-aware adaptation | ProDA · SYNTHIA → Cityscapes | mIoU: 51.9 -> 54.5 | Improved |
| Add depth estimation to image-only adversarial adaptation (ablation) | ProDA adaptation · SYNTHIA → Cityscapes | mIoU: 40.9 -> 41.2 | Improved |
| Add the semantic-geometric consistency loss (ablation) | GAA self-training · SYNTHIA → Cityscapes | mIoU: 51.9 -> 52.3 | Improved |
| Add the geometric adaptation module (ablation) | GAA self-training · SYNTHIA → Cityscapes | mIoU: 51.9 -> 52.9 | Improved |
| Use weighted cross-entropy loss (ablation) | GAA self-training · SYNTHIA → Cityscapes | mIoU: 51.9 -> 53.1 | Improved |
| Combine self-training, GA, WCE, and SGC (ablation) | GAA · SYNTHIA → Cityscapes | mIoU: 51.9 -> 54.5 | Improved |
Closest papers by meaning
- Geometric Unsupervised Domain Adaptation for Semantic Segmentation (ICCV 2021 · similarity 0.90)
- Domain Adaptive Semantic Segmentation With Self-Supervised Depth Estimation (ICCV 2021 · similarity 0.89)
- DODA: Data-Oriented Sim-to-Real Domain Adaptation for 3D Semantic Segmentation (ECCV 2022 · similarity 0.89)
- Learning Semantic Segmentation From Synthetic Data: A Geometrically Guided Input-Output Adaptation Approach (CVPR 2019 · similarity 0.89)
- Open-Set Domain Adaptation for Semantic Segmentation (CVPR 2024 · similarity 0.88)
- Density-guided Translator Boosts Synthetic-to-Real Unsupervised Domain Adaptive Segmentation of 3D Point Clouds (CVPR 2024 · 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.
