Multiple Instance Learning: A Survey of Problem Characteristics and Applications

Abstract

Multiple instance learning (MIL) is a form of weakly supervised learning where training instances are arranged in sets, called bags, and a label is provided for the entire bag. This formulation is gaining interest because it naturally fits various problems and allows to leverage weakly labeled data. Consequently, it has been used in diverse application fields such as computer vision and document classification. However, learning from bags raises important challenges that are unique to MIL. This paper provides a comprehensive survey of the characteristics which define and differentiate the types of MIL problems. Until now, these problem characteristics have not been formally identified and described. As a result, the variations in performance of MIL algorithms from one data set to another are difficult to explain. In this paper, MIL problem characteristics are grouped into four broad categories: the composition of the bags, the types of data distribution, the ambiguity of instance labels, and the task to be performed. Methods specialized to address each category are reviewed. Then, the extent to which these characteristics manifest themselves in key MIL application areas are described. Finally, experiments are conducted to compare the performance of 16 state-of-the-art MIL methods on selected problem characteristics. This paper provides insight on how the problem characteristics affect MIL algorithms, recommendations for future benchmarking and promising avenues for research.

References in corpus (5)

• Parameterized Machine Learning for High-Energy Physics
• Simultaneous Detection and Segmentation
• Feature Learning from Spectrograms for Assessment of Personality Traits
• Label Stability in Multiple Instance Learning
• Deep Multi-Instance Transfer Learning

Cited by in corpus (63)

• Attention-based Deep Multiple Instance Learning
• Hopfield Networks is All You Need
• Detecting Parkinsonian Tremor from IMU Data Collected In-The-Wild using Deep Multiple-Instance Learning
• Appropriateness of Performance Indices for Imbalanced Data Classification: An Analysis
• MRI-based Alzheimer's disease prediction via distilling the knowledge in multi-modal data
• Weakly Supervised Object Detection in Artworks
• WeGleNet: A Weakly-Supervised Convolutional Neural Network for the Semantic Segmentation of Gleason Grades in Prostate Histology Images
• AdvMIL: Adversarial Multiple Instance Learning for the Survival Analysis on Whole-Slide Images
• Deep Weakly-Supervised Learning Methods for Classification and Localization in Histology Images: A Survey
• Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning
• Bag-Level Aggregation for Multiple Instance Active Learning in Instance Classification Problems
• Automated HER2 Scoring in Breast Cancer Images Using Deep Learning and Pyramid Sampling
• DSCA: A Dual-Stream Network with Cross-Attention on Whole-Slide Image Pyramids for Cancer Prognosis
• Weakly Supervised Attention-based Models Using Activation Maps for Citrus Mite and Insect Pest Classification
• Weakly-Supervised Action Localization with Expectation-Maximization Multi-Instance Learning
• Multiple instance learning on deep features for weakly supervised object detection with extreme domain shifts
• Benchmarking Jetson Edge Devices with an End-to-end Video-based Anomaly Detection System
• ProtoDiv: Prototype-guided Division of Consistent Pseudo-bags for Whole-slide Image Classification
• Introducing instance label correlation in multiple instance learning. Application to cancer detection on histopathological images
• Histopathological Image Classification based on Self-Supervised Vision Transformer and Weak Labels
• From Weakly Supervised Learning to Biquality Learning: an Introduction
• Multiple Instance Neuroimage Transformer
• Action Selection Learning for Multi-label Multi-view Action Recognition
• Weakly Supervised AUC Optimization: A Unified Partial AUC Approach
• Hyperbolic Secant representation of the logistic function: Application to probabilistic Multiple Instance Learning for CT intracranial hemorrhage detection
• Psychophysiological Arousal in Young Children Who Stutter: An Interpretable AI Approach
• Using Multiple Instance Learning for Explainable Solar Flare Prediction
• A Multi-resolution Model for Histopathology Image Classification and Localization with Multiple Instance Learning
• Multiple Instance Learning for ECG Risk Stratification
• Attentive Weakly Supervised land cover mapping for object-based satellite image time series data with spatial interpretation
• MINI-Net: Multiple Instance Ranking Network for Video Highlight Detection
• On the Relevance of Temporal Features for Medical Ultrasound Video Recognition
• HopGAT: Hop-aware Supervision Graph Attention Networks for Sparsely Labeled Graphs
• Using Multiple Instance Learning to Build Multimodal Representations
• Reducing Label Dependency in Human Activity Recognition with Wearables: From Supervised Learning to Novel Weakly Self-Supervised Approaches
• Importance Reweighting for Biquality Learning
• Information-Geometric Set Embeddings (IGSE): From Sets to Probability Distributions
• AMI-Net+: A Novel Multi-Instance Neural Network for Medical Diagnosis from Incomplete and Imbalanced Data
• A bag-to-class divergence approach to multiple-instance learning
• Beyond accuracy: quantifying the reliability of Multiple Instance Learning for Whole Slide Image classification
• Learning from Aggregate Observations
• A Multiclass Multiple Instance Learning Method with Exact Likelihood
• Weakly Supervised Learning Meets Ride-Sharing User Experience Enhancement
• MICACL: Multi-Instance Category-Aware Contrastive Learning for Long-Tailed Dynamic Facial Expression Recognition
• Theory and Algorithms for Shapelet-based Multiple-Instance Learning
• Kernel Self-Attention in Deep Multiple Instance Learning
• Multi-Class Multiple Instance Learning for Predicting Precursors to Aviation Safety Events
• Accounting for Dependencies in Deep Learning Based Multiple Instance Learning for Whole Slide Imaging
• Weakly Supervised Action Selection Learning in Video
• Revisiting Data Complexity Metrics Based on Morphology for Overlap and Imbalance: Snapshot, New Overlap Number of Balls Metrics and Singular Problems Prospect
• Training image classifiers using Semi-Weak Label Data
• Improving Robustness of Malware Classifiers using Adversarial Strings Generated from Perturbed Latent Representations
• Dual Normalization Multitasking for Audio-Visual Sounding Object Localization
• mil-benchmarks: Standardized Evaluation of Deep Multiple-Instance Learning Techniques
• Probabilistic HIV Recency Classification -- A Logistic Regression without Labeled Individual Level Training Data
• A Visual Mining Approach to Improved Multiple-Instance Learning
• Learning with partially separable data
• Graph Convolution Neural Network For Weakly Supervised Abnormality Localization In Long Capsule Endoscopy Videos
• Inverse Graph Identification: Can We Identify Node Labels Given Graph Labels?
• Anomaly Detection with Inexact Labels
• In Defense of LSTMs for Addressing Multiple Instance Learning Problems
• META${}^2$: Memory-efficient taxonomic classification and abundance estimation for metagenomics with deep learning
• Multi Instance Learning For Unbalanced Data