Domain Agnostic Learning with Disentangled Representations

Abstract

Unsupervised model transfer has the potential to greatly improve the generalizability of deep models to novel domains. Yet the current literature assumes that the separation of target data into distinct domains is known as a priori. In this paper, we propose the task of Domain-Agnostic Learning (DAL): How to transfer knowledge from a labeled source domain to unlabeled data from arbitrary target domains? To tackle this problem, we devise a novel Deep Adversarial Disentangled Autoencoder (DADA) capable of disentangling domain-specific features from class identity. We demonstrate experimentally that when the target domain labels are unknown, DADA leads to state-of-the-art performance on several image classification datasets.

Cited by in corpus (35)

• A Comprehensive Survey on Transfer Learning
• Federated Adversarial Domain Adaptation
• Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation
• Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentation
• AADG: Automatic Augmentation for Domain Generalization on Retinal Image Segmentation
• Learning domain-agnostic visual representation for computational pathology using medically-irrelevant style transfer augmentation
• Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift
• Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replay
• Minimum Class Confusion for Versatile Domain Adaptation
• Knowledge Distillation Methods for Efficient Unsupervised Adaptation Across Multiple Domains
• A Review of Single-Source Deep Unsupervised Visual Domain Adaptation
• Review-Based Domain Disentanglement without Duplicate Users or Contexts for Cross-Domain Recommendation
• Learnware: Small Models Do Big
• Robust and Generalizable Visual Representation Learning via Random Convolutions
• Variational Disentanglement for Domain Generalization
• Semantic Segmentation in Multiple Adverse Weather Conditions with Domain Knowledge Retention
• Gradient Regularized Contrastive Learning for Continual Domain Adaptation
• Learning Cross-domain Generalizable Features by Representation Disentanglement
• Class-imbalanced Domain Adaptation: An Empirical Odyssey
• ToAlign: Task-oriented Alignment for Unsupervised Domain Adaptation
• AD-GAN: End-to-end Unsupervised Nuclei Segmentation with Aligned Disentangling Training
• Multi-Target Domain Adaptation via Unsupervised Domain Classification for Weather Invariant Object Detection
• An Image is Worth More Than a Thousand Words: Towards Disentanglement in the Wild
• Prototypical Cross-domain Knowledge Transfer for Cervical Dysplasia Visual Inspection
• Learning Disentangled Representations with Latent Variation Predictability
• Disentanglement Then Reconstruction: Learning Compact Features for Unsupervised Domain Adaptation
• Open Compound Domain Adaptation
• Unified cross-modality feature disentangler for unsupervised multi-domain MRI abdomen organs segmentation
• Interventional Domain Adaptation
• Adversarial Purification through Representation Disentanglement
• Domain-Agnostic Learning with Anatomy-Consistent Embedding for Cross-Modality Liver Segmentation
• Low-Resource Dialogue Summarization with Domain-Agnostic Multi-Source Pretraining
• Mutual Information-based Disentangled Neural Networks for Classifying Unseen Categories in Different Domains: Application to Fetal Ultrasound Imaging
• Domain Agnostic Learning for Unbiased Authentication
• Enhanced Separable Disentanglement for Unsupervised Domain Adaptation