Research map / Label-efficient and robust learning / Domain adaptation and generalization
Stronger, Steadier & Superior: Geometric Consistency in Depth VFM Forges Domain Generalized Semantic Segmentation
Siyu Chen, Ting Han, Changshe Zhang, Xin Luo, Meiliu Wu, Guorong Cai, Jinhe Su
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 depth-aware tokens, depth awareness, attention optimization and DRD (full DepthForge) to REIN (ablation) | REIN (DINOv2-L) · Cityscapes->ACDC Snow | mIoU: 70.6 -> 75.4 | Improved |
| concatenate depth features f_d to REIN (ablation) | REIN (DINOv2-L) · Cityscapes->ACDC Snow | mIoU: 70.6 -> 70.8 | No clear effect |
| use Prompt Depth Anything instead of Depth Anything V2 (ablation) | DepthForge · Cityscapes->ACDC Snow (DINOv2-L) | mIoU: 71.5 -> 70.4 | Hurt |
| add Depth Anything V2 features (ablation) | REIN (DINOv2-L) · Cityscapes->ACDC Snow | mIoU: 70.6 -> 71.5 | Improved |
| add Depth Refinement Decoder (ablation) | REIN + Depth Anything V2 · Cityscapes->ACDC Snow | mIoU: 71.5 -> 72.0 | Improved |
| add depth awareness and attention optimization (ablation) | REIN + DRD · Cityscapes->ACDC Snow | mIoU: 72.0 -> 72.2 | Improved |
| use depth-aware learnable tokens (full design) instead of depth token design (ablation) | DepthForge · Cityscapes->ACDC (avg gain) | average gain: 1.6 -> 3.5 | Improved |
| concatenate depth to input instead of no depth (ablation) | REIN · Cityscapes->ACDC (avg gain) | average gain: -1.1 | Hurt |
| add scale factor (ablation) | DepthForge · Cityscapes->ACDC (avg gain) | average gain: 2.8 -> 3.5 | Improved |
| apply DepthForge on EVA02-L backbone | REIN (EVA02-L) · Cityscapes->ACDC Snow | mIoU: 66.0 -> 69.5 | Improved |
| DepthForge vs FADA on GTA5->Cityscapes | FADA · GTA5->Cityscapes | mIoU: 68.23 -> 69.04 | Improved |
| DepthForge vs FADA on Mapillary (Cityscapes source) | FADA · Cityscapes->Mapillary | mIoU: 75.86 -> 75.93 | No clear effect |
Closest papers by meaning
- Leveraging Depth and Language for Open-Vocabulary Domain-Generalized Semantic Segmentation (NeurIPS 2025 · similarity 0.89)
- Unleashing the Power of Visual Foundation Models for Generalizable Semantic Segmentation (AAAI 2025 · similarity 0.89)
- Stronger Fewer & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · similarity 0.88)
- Collaborating Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · similarity 0.87)
- FisherTune: Fisher-Guided Robust Tuning of Vision Foundation Models for Domain Generalized Segmentation (CVPR 2025 · similarity 0.87)
- UniDSeg: Unified Cross-Domain 3D Semantic Segmentation via Visual Foundation Models Prior (NeurIPS 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.
