Taxonomy Adaptive Cross-Domain Adaptation in Medical Imaging via Optimization Trajectory Distillation

arXiv 2307.14709 ↗First posted 27 Jul 2023cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

The success of automated medical image analysis depends on large-scale and expert-annotated training sets. Unsupervised domain adaptation (UDA) has been raised as a promising approach to alleviate the burden of labeled data collection. However, they generally operate under the closed-set adaptation setting assuming an identical label set between the source and target domains, which is over-restrictive in clinical practice where new classes commonly exist across datasets due to taxonomic inconsistency. While several methods have been presented to tackle both domain shifts and incoherent label sets, none of them take into account the common characteristics of the two issues and consider the learning dynamics along network training. In this work, we propose optimization trajectory distillation, a unified approach to address the two technical challenges from a new perspective. It exploits the low-rank nature of gradient space and devises a dual-stream distillation algorithm to regularize the learning dynamics of insufficiently annotated domain and classes with the external guidance obtained from reliable sources. Our approach resolves the issue of inadequate navigation along network optimization, which is the major obstacle in the taxonomy adaptive cross-domain adaptation scenario. We evaluate the proposed method extensively on several tasks towards various endpoints with clinical and open-world significance. The results demonstrate its effectiveness and improvements over previous methods.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 46 of its 72 references that are in the map. First posted on arXiv 27 Jul 2023.

Builds on

46 of 72 references.

References · 46

2015Domain-Adversarial Training of Neural Networkscited by 3,023
2017Prototypical Networks for Few-shot Learningcited by 3,001
2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
2018Neural Tangent Kernel: Convergence and Generalization in Neural Networkscited by 2,597
2016V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentationcited by 2,284
2016On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minimacited by 1,337
2020Gradient Surgery for Multi-Task Learningcited by 1,210
2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,194
2017Deep Hashing Network for Unsupervised Domain Adaptationcited by 1,181
2017GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networkscited by 816
2018Multi-Task Learning as Multi-Objective Optimizationcited by 809
2021SWAD: Domain Generalization by Seeking Flat Minimacited by 431

34 more

2018HoVer-Net: Simultaneous Segmentation and Classification of Nuclei in Multi-Tissue Histology Imagescited by 387
2016Entropy-SGD: Biasing Gradient Descent Into Wide Valleyscited by 357
2019Semi-supervised Domain Adaptation via Minimax Entropycited by 333
2017Three Factors Influencing Minima in SGDcited by 270
2021Gradient Matching for Domain Generalizationcited by 258
2018Characterizing and Avoiding Negative Transfercited by 226
2020Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identificationcited by 214
2020Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformationcited by 183
2017Generalization and Equilibrium in Generative Adversarial Nets (GANs)cited by 165
2019Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentationcited by 151
2019Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain Adaptation for Medical Image Segmentationcited by 121
2020COVID-19 Image Data Collection: Prospective Predictions Are the Futurecited by 100
2022Improving Generalization in Federated Learning by Seeking Flat Minimacited by 98
2021Lizard: A Large-Scale Dataset for Colonic Nuclear Instance Segmentation and Classificationcited by 89
2021Cross-domain Few-shot Learning with Task-specific Adapterscited by 79
2019BackPACK: Packing more into backpropcited by 77
2020Self-training for Few-shot Transfer Across Extreme Task Differencescited by 75
2021Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptationcited by 67
2021Domain Generalization via Gradient Surgerycited by 60
2019Unsupervised Domain Adaptation via Structured Prediction Based Selective Pseudo-Labelingcited by 53
2021Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencodercited by 51
2017Label Efficient Learning of Transferable Representations across Domains and Taskscited by 49
2021Dynamic Distillation Network for Cross-Domain Few-Shot Recognition with Unlabeled Datacited by 49
2020Unsupervised Instance Segmentation in Microscopy Images via Panoptic Domain Adaptation and Task Re-weightingcited by 38
2019Hessian based analysis of SGD for Deep Nets: Dynamics and Generalizationcited by 37
2020Class-Incremental Domain Adaptationcited by 36
2022Adversarial Feature Augmentation for Cross-domain Few-shot Classificationcited by 36
2022Channel Importance Matters in Few-Shot Image Classificationcited by 20
2021Class-Incremental Domain Adaptation with Smoothing and Calibration for Surgical Report Generationcited by 18
2022Recycling Model Updates in Federated Learning: Are Gradient Subspaces Low-Rank?cited by 16
2022Domain Adaptive Nuclei Instance Segmentation and Classification via Category-aware Feature Alignment and Pseudo-labellingcited by 16
2022Decompose to Adapt: Cross-domain Object Detection via Feature Disentanglementcited by 15
2021TACS: Taxonomy Adaptive Cross-Domain Semantic Segmentationcited by 5
2022Cross-Domain Cross-Set Few-Shot Learning via Learning Compact and Aligned Representationscited by 5

Built on by

Cited by year

Citing papers per year: 2024: 6.

6 papers.