Research map / Label-efficient and robust learning / Domain adaptation and generalization
Hierarchical Spatial and Channel Aggregation for Cross-domain Few-shot Segmentation
Sujun Sun, Mingwu Ren, Haofeng Zhang
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 HSA to the baseline (ablation) | SSP-based baseline · CD-FSS average 1-shot | mean-IoU: 66.53 -> 68.43 | Improved |
| add HCA to the baseline (ablation) | SSP-based baseline · CD-FSS average 1-shot | mean-IoU: 66.53 -> 67.90 | Improved |
| add both HSA and HCA (ablation) | SSP-based baseline · CD-FSS average 1-shot | mean-IoU: 66.53 -> 69.22 | Improved |
| add OPSB on top of HSA and HCA (ablation) | DHANet · CD-FSS average 1-shot | mean-IoU: 69.22 -> 70.93 | Improved |
| add OPSB (ablation) | HSA+HCA model · Chest X-ray 1-shot | mean-IoU: 86.52 -> 86.06 | Hurt |
| add slot attention in HSA and HCA (ablation) | DHANet · CD-FSS average 1-shot | mean-IoU: 67.39 -> 69.22 | Improved |
| add cross attention in slots (ablation) | DHANet · CD-FSS average 1-shot | mean-IoU: 67.81 -> 69.22 | Improved |
| add orthogonality loss (ablation) | DHANet · CD-FSS average 1-shot | mean-IoU: 68.33 -> 69.22 | Improved |
| increase HSA stages from 2 to 3 (ablation) | HSA · CD-FSS average 1-shot | mean-IoU: 68.43 -> 68.44 | No clear effect |
| increase HSA stages from 3 to 4 (ablation) | HSA · CD-FSS average 1-shot | mean-IoU: 68.44 -> 68.11 | Hurt |
| increase HCA stages from 2 to 4 (ablation) | HCA · CD-FSS average 1-shot | mean-IoU: 69.22 -> 68.93 | Hurt |
| use probabilistic prototype bank instead of deterministic prototype (ablation) | OPSB · CD-FSS average 1-shot | mean-IoU: 69.95 -> 70.93 | Improved |
Closest papers by meaning
- Dual-Agent Optimization framework for Cross-Domain Few-Shot Segmentation (CVPR 2025 · similarity 0.91)
- Adapt Before Comparison: A New Perspective on Cross-Domain Few-Shot Segmentation (CVPR 2024 · similarity 0.90)
- Lightweight Frequency Masker for Cross-Domain Few-Shot Semantic Segmentation (NeurIPS 2024 · similarity 0.89)
- Cross-Domain Few-Shot Segmentation via Iterative Support-Query Correspondence Mining (CVPR 2024 · similarity 0.89)
- The Devil is in Low-Level Features for Cross-Domain Few-Shot Segmentation (CVPR 2025 · similarity 0.89)
- Adapter Naturally Serves as Decoupler for Cross-Domain Few-Shot Semantic Segmentation (ICML 2025 · 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.
