Research map / Detection and segmentation / CNN features
CDNet: Centripetal Direction Network for Nuclear Instance Segmentation
Hongliang He, Zhongyi Huang, Yao Ding, Guoli Song, Lin Wang, Qian Ren, Pengxu Wei, Zhiqiang Gao, Jie Chen
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 direction-class supervision (k=8) (ablation) | U-Net · MoNuSeg | AJI: 0.5910 -> 0.6095 | Improved |
| add direction-class supervision (k=8) (ablation) | U-Net · CPM17 | AJI: 0.6648 -> 0.7101 | Improved |
| increase direction classes from 8 to 16 (ablation) | U-Net · MoNuSeg | AJI: 0.6095 -> 0.6003 | No clear effect |
| increase direction classes from 8 to 16 (ablation) | U-Net · CPM17 | AJI: 0.7101 -> 0.7047 | No clear effect |
| add center point detection task to DGM (ablation) | U-Net+DGM · MoNuSeg | AJI: 0.6095 -> 0.6153 | Improved |
| add center point detection task to DGM (ablation) | U-Net+DGM · CPM17 | AJI: 0.7101 -> 0.7173 | Improved |
| add interaction between supervision branches in DGM (ablation) | U-Net+DGM · MoNuSeg | AJI: 0.6153 -> 0.6196 | Improved |
| add interaction between supervision branches in DGM (ablation) | U-Net+DGM · CPM17 | AJI: 0.7173 -> 0.7232 | Improved |
| add direction difference map (DDM) at inference (ablation) | CDNet · MoNuSeg | AJI: 0.6196 -> 0.6331 | Improved |
| add direction difference map (DDM) at inference (ablation) | CDNet · CPM17 | AJI: 0.7232 -> 0.7326 | Improved |
| add direction difference map (DDM) at inference (ablation) | CDNet · MoNuSeg | F1: 0.8708 -> 0.8705 | No clear effect |
| add direction-class supervision (k=8) (ablation) | U-Net · CPM17 | F1: 0.8688 -> 0.9167 | Improved |
Later papers that compared with this method
2 later accepted papers report CDNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Mudslide: A Universal Nuclear Instance Segmentation Method (CVPR 2024 · better on 2 of 2 shared results)
- FocusCut: Diving Into a Focus View in Interactive Segmentation (CVPR 2022 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Mudslide: A Universal Nuclear Instance Segmentation Method (CVPR 2024 · similarity 0.86)
- Supervise Less, See More: Training-free Nuclear Instance Segmentation with Prototype-Guided Prompting (ICML 2026 · similarity 0.86)
- Unleashing the Power of Prompt-driven Nucleus Instance Segmentation (ECCV 2024 · similarity 0.86)
- RepSNet: A Nucleus Instance Segmentation Model Based on Boundary Regression and Structural Re-Parameterization (IJCV 2025 · similarity 0.85)
- Are nuclear masks all you need for improved out-of-domain generalisation? A closer look at cancer classification in histopathology (NeurIPS 2024 · similarity 0.81)
- FuseUNet: A Multi-Scale Feature Fusion Method for U-like Networks (ICML 2025 · similarity 0.81)
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.
