Research map / Vision-language and multimodal / CLIP and zero-shot
Bi-VLGM: Bi-Level Class-Severity-Aware Vision-Language Graph Matching for Text Guided Medical Image Segmentation
Wenting Chen, Jie Liu, Tianming Liu, Yixuan Yuan
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 word-level VLGM (ablation) | HRNetV2 baseline · IDRiD | mIoU: 47.74 -> 50.00 | Improved |
| add sentence-level VLGM (ablation) | HRNetV2 baseline · IDRiD | mIoU: 47.74 -> 50.50 | Improved |
| add sentence-level VLGM on top of word-level (ablation) | HRNetV2 + word-level VLGM · IDRiD | mIoU: 50.00 -> 51.37 | Improved |
| add word-level VLGM on top of sentence-level (ablation) | HRNetV2 + sentence-level VLGM · IDRiD | mIoU: 50.50 -> 51.37 | Improved |
| use five severity prompt templates instead of one (ablation) | Bi-VLGM · IDRiD | mIoU: 50.51 -> 51.37 | Improved |
| use two templates instead of one (ablation) | Bi-VLGM · IDRiD | mIoU: 50.51 -> 50.90 | Improved |
| use graph matching instead of contrastive loss (ablation) | Bi-VLGM · IDRiD | mIoU: 49.46 -> 51.37 | Improved |
| use graph matching instead of contrastive loss (ablation) | Bi-VLGM · IDRiD | F1: 65.08 -> 66.71 | Improved |
| use graph matching instead of contrastive loss (ablation) | Bi-VLGM · IDRiD | AUPR: 66.23 -> 68.12 | Improved |
| add word-level VLGM (ablation) | HRNetV2 baseline · IDRiD | SE class mIoU: 51.04 -> 59.53 | Improved |
| add word-level VLGM (ablation) | HRNetV2 baseline · IDRiD | MA IoU: 28.72 -> 27.66 | Hurt |
| Bi-VLGM vs best baseline M2MRF | Bi-VLGM · DDR | mIoU: 31.78 vs 30.09 | Improved |
Closest papers by meaning
- Enhancing Medical Large Vision-Language Models via Alignment Distillation (AAAI 2026 · similarity 0.86)
- DuSSS: Dual Semantic Similarity-Supervised Vision-Language Model for Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.86)
- BiomedCoOp: Learning to Prompt for Biomedical Vision-Language Models (CVPR 2025 · similarity 0.86)
- GMAI-VL & GMAI-VL-5.5M: A Large Vision-Language Model and a Comprehensive Multimodal Dataset Towards General Medical AI (AAAI 2026 · similarity 0.85)
- LLaDA-MedV: Exploring Large Language Diffusion Models for Biomedical Image Understanding (CVPR 2026 · similarity 0.85)
- ExGra-Med: Extended Context Graph Alignment for Medical Vision-Language Models (NeurIPS 2025 · similarity 0.85)
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.
