Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning To Relate Depth and Semantics for Unsupervised Domain Adaptation
Suman Saha, Anton Obukhov, Danda Pani Paudel, Menelaos Kanakis, Yuhua Chen, Stamatios Georgoulis, Luc Van Gool
Check 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 |
|---|---|---|---|
| Use CTRL to align a joint space of task-specific and cross-task features | Semantic and depth UDA model · SYNTHIA → Cityscapes (7 classes, full resolution, EP2) | mIoU: 69.2 to 70.7 | Improved |
| Use CTRL to align a joint space of task-specific and cross-task features | Semantic and depth UDA model · SYNTHIA → Mapillary (7 classes, low resolution, EP3) | mIoU: +2.9% | Improved |
| Add depth supervision to semantic supervision (ablation) | C1 semantic-only model · SYNTHIA → Cityscapes validation | mIoU: 30.7 -> 35.2 | Improved |
| Add SRH supervision to semantic and depth supervision in supervised-only training (ablation) | C2 · SYNTHIA → Cityscapes validation | mIoU: 35.2 -> 33.1 | Hurt |
| Add depth-space adversarial alignment to semantic-space alignment (ablation) | C5 · SYNTHIA → Cityscapes validation | mIoU: 40.8 -> 40.2 | Hurt |
| Add iterative self-learning to the CTRL configuration (ablation) | C8 · SYNTHIA → Cityscapes validation | mIoU: 42.1 -> 44.9 | Improved |
| Add SRH to iterative self-learning with semantic and depth supervision and adversarial alignment (ablation) | C10 · SYNTHIA → Cityscapes validation | mIoU: 42.8 -> 44.9 | Improved |
| Use CTRL and SRH rather than DADA's approach | Monocular depth estimation model · Cityscapes validation (trained on SYNTHIA → Cityscapes, EP1) | |Rel|: 0.6 to 0.3 | Improved |
| Use the proposed method rather than DADA | Semantic UDA model · Virtual KITTI → KITTI | mIoU: 53.0 to 54.6 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DADA — DADA: Depth-Aware Domain Adaptation in Semantic Segmentation (ICCV 2019) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Domain Adaptive Semantic Segmentation With Self-Supervised Depth Estimation (ICCV 2021 · similarity 0.90)
- Learning Across Tasks and Domains (ICCV 2019 · similarity 0.90)
- Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training (ECCV 2018 · similarity 0.90)
- Universal Domain Adaptation for Semantic Segmentation (CVPR 2025 · similarity 0.88)
- Balanced Learning for Domain Adaptive Semantic Segmentation (ICML 2025 · similarity 0.88)
- UniDSeg: Unified Cross-Domain 3D Semantic Segmentation via Visual Foundation Models Prior (NeurIPS 2024 · similarity 0.88)
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.
