![[2602.20187] AINet: Anchor Instances Learning for Regional Heterogeneity in Whole Slide Image](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20187] AINet: Anchor Instances Learning for Regional Heterogeneity in Whole Slide Image
Summary
The paper introduces AINet, a novel framework for whole slide image analysis that addresses regional heterogeneity through anchor instances, enhancing multi-instance learning performance.
Why It Matters
This research is significant as it tackles the challenges of tumor sparsity and morphological diversity in whole slide images, which are critical for accurate medical diagnoses. The proposed AINet framework improves efficiency and effectiveness in image analysis, potentially impacting cancer research and diagnostics.
Key Takeaways
- AINet utilizes anchor instances to improve representation in multi-instance learning.
- The dual-level anchor mining (DAM) module selects the most informative instances for analysis.
- The anchor-guided region correction (ARC) module enhances regional representation by leveraging complementary information.
- AINet achieves superior performance with fewer computational resources compared to existing methods.
- Both DAM and ARC can be integrated into current multi-instance learning frameworks.
Electrical Engineering and Systems Science > Image and Video Processing arXiv:2602.20187 (eess) [Submitted on 21 Feb 2026] Title:AINet: Anchor Instances Learning for Regional Heterogeneity in Whole Slide Image Authors:Tingting Zheng, Hongxun Yao, Kui Jiang, Sicheng Zhao, Yi Xiao View a PDF of the paper titled AINet: Anchor Instances Learning for Regional Heterogeneity in Whole Slide Image, by Tingting Zheng and 4 other authors View PDF HTML (experimental) Abstract:Recent advances in multi-instance learning (MIL) have witnessed impressive performance in whole slide image (WSI) analysis. However, the inherent sparsity of tumors and their morphological diversity lead to obvious heterogeneity across regions, posing significant challenges in aggregating high-quality and discriminative representations. To address this, we introduce a novel concept of anchor instance (AI), a compact subset of instances that are representative within their regions (local) and discriminative at the bag (global) level. These AIs act as semantic references to guide interactions across regions, correcting non-discriminative patterns while preserving regional diversity. Specifically, we propose a dual-level anchor mining (DAM) module to \textbf{select} AIs from massive instances, where the most informative AI in each region is extracted by assessing its similarity to both local and global embeddings. Furthermore, to ensure completeness and diversity, we devise an anchor-guided region correction (ARC) m...