Domain Adaptation / Transfer learning / Semi-supervised paper list

List recent papers related to domain adaptation in different type of applications. Some are referred from Awesome Domain Adaptation.

Survey

• Deep Visual Domain Adaptation: A Survey (Neurocomputing'18)
• A Survey on Transfer Learning (IEEE TKDE'09)

Theory

• A General Upper Bound for Unsupervised Domain Adaptation (ArXiv'19.10)
• An analytic theory of generalization dynamics and transfer learning in deep linear networks (ICLR'19)
• A Kernel Two-Sample Test (JMLR'2012): MMD
• A theory of learning from different domains (Machine Learning' 2010): H-divergence

Classification

Unsupervised Domain Adaptation

• Beyond Sharing Weights for Deep Domain Adaptation (PAMI'19)[ArXiv]: MMD
• Cluster Alignment with a Teacher for Unsupervised Domain Adaptation (ICCV'19)
• Unsupervised Domain Adaptation via Regularized Conditional Alignment (ICCV'19)
• Drop to Adapt: Learning Discriminative Features for Unsupervised Domain Adaptation (ICCV'19): adapt classifier
• Bayesian Uncertainty Matching for Unsupervised Domain Adaptation (IJCAI'19): classifier adaptation based
• Attending to Discriminative Certainty for Domain Adaptation (CVPR'19)
• Contrastive Adaptation Network for Unsupervised Domain Adaptation (CVPR'19)
• Sliced Wasserstein Discrepancy for Unsupervised Domain Adaptation (CVPR'19)
• AdaGraph: Unifying Predictive and Continuous Domain Adaptation through Graphs (CVPR'19)
• Domain-Symmetric Networks for Adversarial Domain Adaptation (CVPR'19)
• Transferable Adversarial Training: A General Approach to Adapting Deep Classifiers (ICML'19): adapt classifier
• Domain Agnostic Learning with Disentangled Representations (ICML'19)
• Transferability vs. Discriminability: Batch Spectral Penalization for Adversarial Domain Adaptation (ICML'19)
• Learning What and Where to Transfer (ICML'19)
• Bridging Theory and Algorithm for Domain Adaptation (ICML'19)
• On Learning Invariant Representations for Domain Adaptation (ICML'19)
• Augmented Cyclic Adversarial Learning for Low Resource Domain Adaptation (ICLR'19)
• Regularized Learning for Domain Adaptation under Label Shifts (ICLR'19)
• Improving the Generalization of Adversarial Training with Domain Adaptation (ICLR'19)
• Multi-Domain Adversarial Learning (ICLR'19)
• Learning Factorized Representations for Open-set Domain Adaptation (ICLR'19)
• Transferable attention for domain adaptation (AAAI'19): attention mechanism, adversarial
• A Unified Feature Disentangler for Multi-Domain Image Translation and Manipulation (NIPS'18)
• Unsupervised Image-to-Image Translation Using Domain-Specific Variational Information Bound (NIPS'18): MMD
• Conditional Adversarial Domain Adaptation (NIPS'18)
• Co-regularized Alignment for Unsupervised Domain Adaptation (NIPS'18)
• Graph Adaptive Knowledge Transfer for Unsupervised Domain Adaptation (ECCV'18)
• Deep Adversarial Attention Alignment for Unsupervised Domain Adaptation:the Benefit of Target Expectation Maximization (ECCV'18)
• DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptation (ECCV'18): Optimal transport, Wasserstein distance
• Maximum Classifier Discrepancy for Unsupervised Domain Adaptation (CVPR'18 Oral): Discriminative domain invariant feature, Ensemble-based DA
• Detach and Adapt: Learning Cross-Domain Disentangled Deep Representation (CVPR'18 Spotlight)
• Generate To Adapt: Aligning Domains using Generative Adversarial Networks (CVPR'18 Spotlight)
• Collaborative and Adversarial Network for Unsupervised domain adaptation (CVPR'18): Adversarial-based DA
• Unsupervised Domain Adaptation with Similarity Learning (CVPR'18)
• Duplex Generative Adversarial Network for Unsupervised Domain Adaptation (CVPR'18)
• Learning Semantic Representations for Unsupervised Domain Adaptation (ICML'18)
• Adversarial Dropout Regularization (ICLR'18): Discriminative domain invariant feature, Ensemble-based DA
• Self-ensembling for visual domain adaptation (ICLR'18): Ensemble-based DA
• Adaptive Batch Normalization for practical domain adaptation (Pattern Recognition'18): normalized-based method
• Multi-Adversarial Domain Adaptation (AAAI'18 Oral): class-aware domain discrepancy, discriminative domain invariant feature
• Wasserstein Distance Guided Representation Learning for Domain Adaptation (AAAI'18)
• Joint distribution optimal transportation for domainadaptation (NIPS'17): Optimal transport, Wasserstein distance
• AutoDIAL: Automatic DomaIn Alignment Layers (ICCV'17): batch normalization layer
• Associative Domain Adaptation (ICCV'17): Discriminative domain invariant feature
• Adversarial discriminative domain adaptation (CVPR'17)
• Unsupervised Pixel–Level Domain Adaptation with Generative Adversarial Networks (CVPR'17)
• Mind the Class Weight Bias: Weighted Maximum Mean Discrepancyfor Unsupervised Domain Adaptation (CVPR'17): MMD based method
• Deep Transfer Learning with Joint Adaptation Networks (ICML'17): JMMD
• Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning (ICLR'17): MMD based method
• Revisiting Batch Normalization For Practical Domain Adaptation (ICLR'17 workshop): normalization-based method
• Correlation Alignment for Unsupervised Domain Adaptation (DA in CV'17): CORAL
• Coupled Generative Adversarial Networks (NIPS'16)
• Learning Transferrable Representations for Unsupervised Domain Adaptation (NIPS'16): Discriminative domain invariant feature
• Unsupervised Domain Adaptation with ResidualTransfer Networks (NIPS'16): MMD based method
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation (ECCV'16 workshop)
• Domain adversarial training of neural networks (JMLR'16)
• Optimal Transport for Domain Adaptation (TPAMI'16): Optimal transport, Wasserstein distance
• Return of Frustratingly Easy Domain Adaptation (AAAI'16): correlation distance
• Simultaneous deep transfer across domains and tasks (CVPR'15)
• Learning transferable features with deep adaptation networks (ICML'15): MMD
• Deep Domain Confusion: Maximizing for Domain Invariance (ArXiv'14): MMD
• Transfer Feature Learning with Joint Distribution Adaptation (ICCV'13): class-aware domain discrepancy

Universal DA

• Universal Domain Adaptation (CVPR'19)

Multi-Source DA

• Moment Matching for Multi-Source Domain Adaptation (ICCV'19 Oral): new generalisation bounds, DomainNet dataset
• Adversarial Multiple Source Domain Adaptation (NIPS'18): new generalisation bounds

Domain Generalization

Labeled source domain data only, no unlabeled/labeled target domain data

• Domain Generalization via Model-Agnostic Learning of Semantic Features (NIPS'19): meta-learning for multi source DG
• Domain Generalization by Solving Jigsaw Puzzles (CVPR'19 Oral)
• MetaReg: Towards Domain Generalization using Meta-Regularization (NIPS'18): meta-learning for multi source DG
• Deep Domain Generalization via Conditional Invariant Adversarial Networks (ECCV'18)
• Learning to Generalize: Meta-Learning for Domain Generalization (AAAI'18): meta-learning for multi source DG

Semi-supervised DA

• Semi-supervised Domain Adaptation via Minimax Entropy (ICCV'19)
• Semi-Supervised Domain Adaptation With Subspace Learning for Visual Recognition (CVPR'15): subspace learning
• Semi-Supervised Domain Adaptation with Instance Constraints (CVPR'13): label smoothing

Open-Set DA

Source label set is a subset of target label set

• Learning Factorized Representations for Open-Set Domain Adaptation (ICLR'19)
• Open Set Domain Adaptation by Backpropagation (ECCV'18)

Unknown classes exist in source and target domain, assume common classes are known

• Open Set Domain Adaptation (ICCV'17)

Partial DA

Target label set is a subset of source label set

• Learning to Transfer Examples for Partial Domain Adaptation (CVPR'19)
• Partial Adversarial Domain Adaptation (ECCV'18)
• Partial Transfer Learning with Selective Adversarial Networks (CVPR'18)
• Importance Weighted Adversarial Nets for Partial Domain Adaptation (CVPR'18)

Weakly Supervised DA

Abundant source domain data, limited target domain data with labels

• Not All Areas Are Equal: Transfer Learning for Semantic Segmentation via Hierarchical Region Selection (CVPR'19 Oral)

Partly shared label sets, some labeled target domain data

• Label efficient learning of transferable representations acrosss domains and tasks (NIPS'17): entropy-based DA

Video Classification

Human action recognition

• Temporal Attentive Alignment for Large-Scale Video Domain Adaptation (ICCV'19 Oral)
• Learning Transferable Self-attentive Representations for Action Recognition in Untrimmed Videos with Weak Supervision (AAAI'19)
• Deep domain adaptation in action space (BMVC'18)
• Dual many-to-one-encoder-based transfer learning for cross-dataset human action recognition (Image and Vision Computing'16)
• Human action recognition acrossdatasets by foreground-weighted histogram decomposition (CVPR'14)

Face Recognition

• Unsupervised Domain Adaptation for Face Recognition in Unlabeled Videos (ICCV'17)

Semantic Segmentation

Unsupervised Domain Adaptation

• Drop to Adapt: Learning Discriminative Features for Unsupervised Domain Adaptation (ICCV'19): adapt classifier
• Guided Curriculum Model Adaptation and Uncertainty-Aware Evaluation for Semantic Nighttime Image Segmentation (ICCV'19)
• DLOW: Domain Flow for Adaptation and Generalization (CVPR'19 Oral)
• ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation (CVPR'19)
• Learning Semantic Segmentation from Synthetic Data: A Geometrically Guided Input-Output Adaptation Approach (CVPR'19)
• Bidirectional Learning for Domain Adaptation of Semantic Segmentation (CVPR'19)
• CrDoCo: Pixel-level Domain Transfer with Cross-Domain Consistency (CVPR'19)
• Taking A Closer Look at Domain Shift: Category-level Adversaries for Semantics Consistent Domain Adaptation (CVPR'19)
• All about Structure: Adapting Structural Information across Domains for Boosting Semantic Segmentation (CVPR'19)
• Weakly Supervised Adversarial Domain Adaptation for Semantic Segmentation in Urban Scenes (TIP)
• SPIGAN: Privileged Adversarial Learning from Simulation (ICLR'19)
• Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training (ECCV'18)
• Penalizing Top Performers: Conservative Loss for Semantic Segmentation Adaptation (ECCV'18)
• Domain transfer through deep activation matching (ECCV'18)
• Cycada: Cycle-consistent adversarial domain adaptation (ICML'18)
• Learning to adapt structured output space for semantic segmentation (CVPR'18 Spotlight)
• Learning from synthetic data: Addressing domain shift for semantic segmentation (CVPR'18 Spotlight)
• ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes (CVPR'18)
• Conditional Generative Adversarial Network for Structured Domain Adaptation (CVPR'18)
• Image to Image Translation for Domain Adaptation (CVPR'18)
• Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes (ICCV'17)
• FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation (ArXiv'16)

Instance Segmentation

• SRDA: Generating Instance Segmentation Annotation Via Scanning, Reasoning And Domain Adaptation (ECCV'18)

Feature Disentanglement

• Domain Agnostic Learning with Disentangled Representations (ICML'19)
• A Unified Feature Disentangler for Multi-Domain Image Translation and Manipulation (NIPS'18)
• Detach and Adapt: Learning Cross-Domain Disentangled Deep Representation (CVPR'18 Spotlight)

Person Re-identification

• Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-identification (CVPR'19)
• Domain Adaptation through Synthesis for Unsupervised Person Re-identification (ECCV'18)
• Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification (CVPR'18)
• Adaptation and Re-Identification Network: An Unsupervised Deep Transfer Learning Approach to Person Re-Identification (CVPRW'18)

Depth Estimation

• Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation (CVPR'19)
• T2net: Synthetic-to-realistic translation for solving single-image depth estimation tasks (ECCV'18)
• AdaDepth: Unsupervised Content Congruent Adaptation for Depth Estimation (CVPR'18)
• Real-Time Monocular Depth Estimation using Synthetic Data with Domain Adaptation via Image Style Transfer (CVPR'18)

Stereo

• Learning to Adapt for Stereo (CVPR'19)
• Learning Monocular Depth by Distilling Cross-domain Stereo Networks (ECCV'18)
• Unsupervised adaptation for deep stereo (ICCV'17)

Object Detection

• Strong-Weak Distribution Alignment for Adaptive Object Detection (CVPR'19): [Page]
• Automatic adaptation of object detectors to new domains using self-training (CVPR'19)
• AugGAN: Cross Domain Adaptation with GAN-based Data Augmentation (ECCV'18)
• Domain Adaptive Faster R-CNN for Object Detection in the Wild (CVPR'18)

Reinforcement Learning

• Sim-Real Joint Reinforcement Transfer for 3D Indoor Navigation (CVPR'19)

Multi-Domain

• Multi-Domain Adversarial Learning (ICLR'19)

Zero-shot Learning

• Zero-Shot Deep Domain Adaptation (ECCV'18)

Other

Scene Text Detection and Recognition

• GA-DAN: Geometry-Aware Domain Adaptation Network for Scene Text Detection and Recognition (ICCV'19)

Adversarial Attack

• Improving the Generalization of Adversarial Training with Domain Adaptation (ICLR'19)

Brain-Machine Interfaces

• Adversarial Domain Adaptation for Stable Brain-Machine Interfaces (ICLR'19)

Metric Learning

• Unsupervised Domain Adaptation for Distance Metric Learning (ICLR'19)

3D Keypoint Estimation

• Unsupervised Domain Adaptation for 3D Keypoint Estimation via View Consistency (ECCV'18)

Museum Exhibit Identification

• Museum Exhibit Identification Challenge for the Supervised Domain Adaptation and Beyond (ECCV'18)

Autonomous Driving

• Real-to-Virtual Domain Unification for End-to-End Autonomous Driving (ECCV'18)

Gaze estimation and human hand pose estimation

• Learning from Simulated and Unsupervised Images through Adversarial Training (CVPR'17)

Semi-supervised Learning

Classification

• Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning (TPAMI'18)
• Realistic Evaluation of Deep Semi-Supervised Learning Algorithms (NIPS'18)
• mixup: Beyond Empirical Risk Minimization (ICLR'18)
• Semi-Supervised Learning with Generative Adversarial Networks (ICML'16 workshop)
• Frustratingly Easy Domain Adaptation (ACL'07): regularization
• Semi-supervised Learning by Entropy Minimization (NIPS'04)

Semantic Segmentation

• Universal Semi-Supervised Semantic Segmentation (ICCV'19)
• Adversarial Learning for Semi-Supervised Semantic Segmentation (BMVC'18)

Transfer Learning

Image generation

• Image Generation From Small Datasets via Batch Statistics Adaptation (ICCV'19)