QG-MIL: A Gated Transformer Aggregator for Domain-Agnostic Multiple Instance Learning in Medical Imaging

• URL: http://arxiv.org/abs/2606.20027v1
• Date: Thu, 18 Jun 2026 09:58:50 GMT
• Title: QG-MIL: A Gated Transformer Aggregator for Domain-Agnostic Multiple Instance Learning in Medical Imaging
• Abstract summary: Attention-based Multiple Instance Learning aggregators in medical imaging are prone to attention concentration.<n>We introduce QG-MIL, a gated transformer aggregator that addresses this through four synergistic architectural components.<n>We evaluate QG-MIL across six benchmarks spanning whole-slide pathology and cell-level hematology, covering two fundamentally different MIL scales.
• Score: 2.593919493461823
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Attention-based Multiple Instance Learning aggregators in medical imaging are prone to attention concentration, producing overconfident and unstable predictions. We introduce QG-MIL, a gated transformer aggregator that addresses this through four synergistic architectural components: RMSNorm-based pre-normalization, per-head QK normalization, fine-grained attention output gating, and SwiGLU-style feed-forward modules. Together, these design choices stabilize training and distribute attention more uniformly across instances without auxiliary losses, masking, or multi-stage regularization. We evaluate QG-MIL across six benchmarks spanning whole-slide pathology and cell-level hematology, covering two fundamentally different MIL scales. The best-performing QG-MIL variants outperform leading baselines on all six benchmarks, with an average improvement of +6.1 mean macro F1 points. Attention overlays and attention mass analysis confirm more distributed instance weighting. Ablation studies show that while individual components can match the full model on specific datasets, the QG-MIL design provides the most consistent cross-domain performance and tightest variance when compared to selected baselines. We release a configurable implementation to support reproducibility at: https://github.com/unica-visual-intelligence-lab/QG-MIL

Related papers

• FoundationGeo: Learning Spatial Pixel-Wise Fields for Monocular Metric Geometry [52.232826445242644]
  FoundationGeo bridges relative and metric prediction via spatial calibration and principled data design.<n>Stage 1 learns a high-fidelity, affine-invariant geometry model by initializing with DINOv3.<n>Stage 2 moves beyond global scaling by introducing pixel-wise calibration fields for metric estimation.
  arXiv  Detail & Related papers  (2026-07-13T14:10:01Z)
• ASMIL: Attention-Stabilized Multiple Instance Learning for Whole Slide Imaging [95.36377714370325]
  We show that attention-based multiple instance learning methods exhibit a new failure mode: unstable attention dynamics.<n>We introduce attention-stabilized multiple instance learning (ASMIL), a novel unified framework.<n>ASMIL uses an anchor model to stabilize attention, replaces softmax with a normalized sigmoid function in the anchor to prevent over-concentration, and applies token random dropping to mitigate overfitting.
  arXiv  Detail & Related papers  (2026-03-01T18:31:13Z)
• scDFM: Distributional Flow Matching Model for Robust Single-Cell Perturbation Prediction [12.48933770510505]
  We present scDFM, a generative framework based on conditional flow matching.<n> scDFM aligns perturbed and control populations beyond cell-level correspondences.
  arXiv  Detail & Related papers  (2026-02-06T17:00:21Z)
• From Global to Granular: Revealing IQA Model Performance via Correlation Surface [83.65597122328133]
  We present textbfGranularity-Modulated Correlation (GMC), which provides a structured, fine-grained analysis of IQA performance.<n>GMC includes a textbfDistribution Regulator that regularizes correlations to mitigate biases from non-uniform quality distributions.<n>Experiments on standard benchmarks show that GMC reveals performance characteristics invisible to scalar metrics, offering a more informative and reliable paradigm for analyzing, comparing, and deploying IQA models.
  arXiv  Detail & Related papers  (2026-01-29T13:55:26Z)
• MMSF: Multitask and Multimodal Supervised Framework for WSI Classification and Survival Analysis [8.125488986754968]
  We introduce MMSF, a multitask and multimodal supervised framework built on a linear-complexity MIL backbone.<n>Experiments on CAMELYON16 and TCGA-NSCLC demonstrate 2.1--6.6% accuracy and 2.2--6.9% AUC improvements over competitive baselines.
  arXiv  Detail & Related papers  (2026-01-28T08:05:46Z)
• MS-ISSM: Objective Quality Assessment of Point Clouds Using Multi-scale Implicit Structural Similarity [65.85858856481131]
  unstructured and irregular nature of point clouds poses a significant challenge for objective quality assessment (PCQA)<n>We propose the Multi-scale Implicit Structural Similarity Measurement (MS-ISSM)
  arXiv  Detail & Related papers  (2026-01-03T14:58:52Z)
• GAS-MIL: Group-Aggregative Selection Multi-Instance Learning for Ensemble of Foundation Models in Digital Pathology Image Analysis [6.45975531973783]
  GAS-MIL is a flexible ensemble framework that seamlessly integrates features from multiple foundation models.<n>It achieves superior or on-par performance relative to individual FMs and established MIL methods.<n>It provides a scalable foundation for future multimodal and precision oncology applications.
  arXiv  Detail & Related papers  (2025-10-03T22:59:40Z)
• Single GPU Task Adaptation of Pathology Foundation Models for Whole Slide Image Analysis [8.076987502347327]
  Pathology foundation models (PFMs) have emerged as powerful tools for analyzing whole slide images (WSIs)<n>TAPFM uses vision transformer (vit) attention for MIL aggregation while optimizing both for feature representations and attention weights.<n> evaluated on mutation prediction tasks for bladder cancer and lung adenocarcinoma.
  arXiv  Detail & Related papers  (2025-06-05T15:56:45Z)
• MamMIL: Multiple Instance Learning for Whole Slide Images with State Space Models [56.37780601189795]
  We propose a framework named MamMIL for WSI analysis. We represent each WSI as an undirected graph. To address the problem that Mamba can only process 1D sequences, we propose a topology-aware scanning mechanism.
  arXiv  Detail & Related papers  (2024-03-08T09:02:13Z)
• Multiplex-detection Based Multiple Instance Learning Network for Whole Slide Image Classification [2.61155594652503]
  Multiple instance learning (MIL) is a powerful approach to classify whole slide images (WSIs) for diagnostic pathology. We propose a novel multiplex-detection-based multiple instance learning (MDMIL) to tackle the issues above. Specifically, MDMIL is constructed by the internal query generation module (IQGM) and the multiplex detection module (MDM)
  arXiv  Detail & Related papers  (2022-08-06T14:36:48Z)
• Local Attention Graph-based Transformer for Multi-target Genetic Alteration Prediction [0.22940141855172028]
  We propose a general-purpose local attention graph-based Transformer for MIL (LA-MIL) We demonstrate that LA-MIL achieves state-of-the-art results in mutation prediction for gastrointestinal cancer. This suggests that local self-attention sufficiently models dependencies on par with global modules.
  arXiv  Detail & Related papers  (2022-05-13T14:24:24Z)
• Self-Guided Adaptation: Progressive Representation Alignment for Domain Adaptive Object Detection [86.69077525494106]
  Unsupervised domain adaptation (UDA) has achieved unprecedented success in improving the cross-domain robustness of object detection models. Existing UDA methods largely ignore the instantaneous data distribution during model learning, which could deteriorate the feature representation given large domain shift. We propose a Self-Guided Adaptation (SGA) model, target at aligning feature representation and transferring object detection models across domains.
  arXiv  Detail & Related papers  (2020-03-19T13:30:45Z)
• Q-GADMM: Quantized Group ADMM for Communication Efficient Decentralized Machine Learning [66.18202188565922]
  We propose a communication-efficient decentralized machine learning (ML) algorithm, coined QGADMM (QGADMM)<n>We develop a novel quantization method to adaptively adjust modelization levels and their probabilities, while proving the convergence of QGADMM for convex functions.
  arXiv  Detail & Related papers  (2019-10-23T10:47:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.