Research map / Vision-language and multimodal / CLIP and zero-shot
IVAAN: Instance-level Vision-Language Alignment via Attribute-Guided Text Prompts Generation for Nuclei Analysis
Jaehoon Jeong, Yi Hu, Soopil Kim, Jongseong Jang, Soonyoung Lee, Sang Hyun Park
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 instance-level VL alignment with fixed class-organ prompts (ablation) | Mask2Former nuclei model · PanNuke | PQ: 57.3 -> 62.3 | Improved |
| add attribute-guided prompts with quantile bins (ablation) | VL alignment with fixed prompts · PanNuke | PQ: 62.3 -> 63.8 | Improved |
| use entropy-based binning instead of quantile bins (ablation) | attribute prompts · PanNuke | PQ: 63.8 -> 65.4 | Improved |
| add class tokens with Semantic Interaction Module (ablation) | IVAAN · PanNuke | PQ: 65.4 -> 66.4 | Improved |
| add image-level feature fusion with text embeddings (ablation) | IVAAN · PanNuke | PQ: 66.4 -> 67.3 | Improved |
| use IVAAN over PromptNucSeg-H | nuclei segmentation · PanNuke | mPQ: 0.5123 -> 0.5459 | Improved |
| use IVAAN over PromptNucSeg-H | nuclei segmentation · PanNuke dead class | PQ: 0.161 -> 0.204 | Improved |
| use IVAAN over PromptNucSeg-H | nuclei segmentation · CPM17 (base size) | AJI: Ours-B 0.729 vs PromptNucSeg-B 0.731 AJI; PQ 0.727 vs 0.726 | Mixed |
| use IVAAN over PromptNucSeg-H | nuclei segmentation · MoNuSeg | AJI: 0.622 -> 0.689 | Improved |
| use IVAAN over PromptNucSeg-H | nuclei segmentation · PanNuke Kidney | bPQ: 0.7115 vs 0.6801 (bPQ Kidney) | Hurt |
Closest papers by meaning
- Incremental Nuclei Segmentation from Histopathological Images via Future-class Awareness and Compatibility-inspired Distillation (CVPR 2024 · similarity 0.87)
- Few-Shot Learning from Gigapixel Images via Hierarchical Vision-Language Alignment and Modeling (NeurIPS 2025 · similarity 0.85)
- UniCell: Universal Cell Nucleus Classification via Prompt Learning (AAAI 2024 · similarity 0.84)
- CPLIP: Zero-Shot Learning for Histopathology with Comprehensive Vision-Language Alignment (CVPR 2024 · similarity 0.83)
- Multi-Resolution Pathology-Language Pre-training Model with Text-Guided Visual Representation (CVPR 2025 · similarity 0.83)
- Weakly Supervised Gland Segmentation with Class Semantic Consistency and Purified Labels Filtration (AAAI 2025 · similarity 0.83)
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.
