Research map / Label-efficient and robust learning / Domain adaptation and generalization
Hierarchical Prompt Injector for Domain Generalization Segmentation
Xin Kun Lin, Ruoyu Guo, Jiaqi Guo, Maurice Pagnucco, Yang Song
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 Region Cue to prompts (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.28 -> 70.62 | Improved |
| add Part description to prompts (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.40 -> 70.62 | Improved |
| add Geometry description to prompts (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.44 -> 70.62 | Improved |
| add Boundary Relation to prompts (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.74 -> 70.62 | Improved |
| use R=3 regional prompts instead of 2 (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.48 -> 70.62 | Improved |
| use R=3 regional prompts instead of 4 (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.69 -> 70.62 | Improved |
| add SAC gating (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.73 -> 70.62 | Improved |
| add SCC gating (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.94 -> 70.62 | Improved |
| add spatial localization loss (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.54 -> 70.62 | Improved |
| add semantic consistency loss (ablation) | HPI (CLIP-L+DINOv2-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 69.91 -> 70.62 | Improved |
| add DINOv2 branch (ablation) | HPI (CLIP-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 60.33 -> 70.62 | Improved |
| add DINOv2 branch (ablation) | HPI (EVA02-L) · G->{C,B,M} (synthetic-to-real) average | mIoU: 66.63 -> 69.39 | Improved |
Closest papers by meaning
- Stronger Fewer & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · similarity 0.88)
- Prompting Diffusion Models for Zero-Shot Instance Segmentation (NeurIPS 2026 · similarity 0.88)
- Visual Prompting for Generalized Few-shot Segmentation: A Multi-scale Approach (CVPR 2024 · similarity 0.88)
- Collaborating Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · similarity 0.87)
- APSeg: Auto-Prompt Network for Cross-Domain Few-Shot Semantic Segmentation (CVPR 2024 · similarity 0.87)
- Leveraging Depth and Language for Open-Vocabulary Domain-Generalized Semantic Segmentation (NeurIPS 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.
