Research map / Detection and segmentation / CNN features
WIMFRIS: WIndow Mamba Fusion and Parameter Efficient Tuning for Referring Image Segmentation
Seunghun Moon, Hyunwoo Yu, Haeuk Lee, Suk-Ju Kang
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 window partitioning (4x4) to WMF instead of plain concatenation (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 76.0 -> 77.3 | Improved |
| use 2x2 instead of 4x4 windows (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 77.3 -> 76.5 | Hurt |
| add Multi-Scale Aligner (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 75.7 -> 76.4 | Improved |
| add Emphasis Parameter on top of MSA (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 76.4 -> 76.4 | No clear effect |
| add Mamba Text Adapter (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 75.7 -> 76.8 | Improved |
| combine MSA, EP and MTA (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 75.7 -> 77.3 | Improved |
| use multi-scale kernels 3-5-7 instead of 1-1-1 (ablation) | RFMixer · RefCOCO val | mIoU: 76.7 -> 77.3 | Improved |
| distribute adapters across layers (1,3,5,7,9,11) rather than concentrating them (ablation) | WIMFRIS-B · RefCOCO val | mIoU: 76.6 -> 77.3 | Improved |
| add HMF neck to ETRIS (vs no neck) (ablation) | ETRIS · RefCOCO val | mIoU: 72.2 -> 74.5 | Improved |
| replace ETRIS neck with our HMF block (ablation) | ETRIS · RefCOCO val | mIoU: 74.5 -> 75.7 | Improved |
| add DETRIS neck (vs no neck) (ablation) | DETRIS · RefCOCO val | mIoU: 74.3 -> 75.8 | Improved |
| replace DETRIS neck with our HMF block (ablation) | DETRIS · RefCOCO val | mIoU: 75.8 -> 76.4 | Improved |
Closest papers by meaning
- Densely Connected Parameter-Efficient Tuning for Referring Image Segmentation (AAAI 2025 · similarity 0.87)
- LQMFormer: Language-aware Query Mask Transformer for Referring Image Segmentation (CVPR 2024 · similarity 0.87)
- Prompt-Driven Referring Image Segmentation with Instance Contrasting (CVPR 2024 · similarity 0.86)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.85)
- Curriculum Point Prompting for Weakly-Supervised Referring Image Segmentation (CVPR 2024 · similarity 0.85)
- Boosting Weakly Supervised Referring Image Segmentation via Progressive Comprehension (NeurIPS 2024 · similarity 0.84)
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.
