kgl-prml / AmazingTransferLearning

Transfer Learning, Domain Adaptation

Resources about Transfer Learning

---

Table of contents

• Papers
• Surveys
• Codes
• Scholars
• Datasets
• Typical methods

---

Papers

ICCV && ECCV:

• 20171222 ICCV 2017 Unified Deep Supervised Domain Adaptation and Generalization | Code

• 20171201 ICCV-17 When Unsupervised Domain Adaptation Meets Tensor Representations

• 201711 ICCV-17 Open set domain adaptation

• ICCV-17 Open set domain adaptation

• ICCV-17 AutoDIAL: Automatic DomaIn Alignment Layers

• ICCV-17 CCSA: Unified Deep Supervised Domain Adaptation and Generalization

• ICCV-17 CycleGAN: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks

• ICCV-17 DualGAN: DualGAN: Unsupervised Dual Learning for Image-to-Image Translation

• ICCV-15 Simultaneous Deep Transfer Across Domains and Tasks

• ICCV-13 Transfer Feature Learning with Joint Distribution Adaptation

• ECCV-16 Deep CORAL: Correlation Alignment for Deep Domain Adaptation

• ECCV-16 DRCN: Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation

CVPR

• CVPR-17 ADDA: Adaptative Discriminative Domain Adaptation

• CVPR-17 Asymmetric Tri-training for Unsupervised Domain Adaptation

• CVPR-14 Transfer Joint Matching for Unsupervised Domain Adaptation

• CVPR-14 CNN features off-the-Shelf: An astounding baseline for recognition

• CVPR-14 Learning and transferring mid-Level image representations using convolutional neural networks

NIPS

• 20171128 NIPS-17 Learning Multiple Tasks with Multilinear Relationship Networks

• 20171126 NIPS-17 Label Efficient Learning of Transferable Representations acrosss Domains and Tasks

• NIPS-17 Learning Multiple Tasks with Multilinear Relationship Networks

• NIPS-17 Label Efficient Learning of Transferable Representations acrosss Domains and Tasks

• NIPS-17 FADA: Few-Shot Adversarial Domain Adaptation

• NIPS-17 JDOT: Joint distribution optimal transportation for domain adaptation

• 20171222 NIPS 2017 Few-Shot Adversarial Domain Adaptation

• 20171226 NIPS 2016 Domain Separation Networks | Code

• NIPS-16 RTN: Unsupervised Domain Adaptation with Residual Transfer Networks

• NIPS-16 DSN: Domain Separation Networks

• NIPS-14 How transferable are features in deep neural networks?

ICML && JMLR

• 2017-ICML Deep Transfer Learning with Joint Adaptation Networks Code

• ICML-17 JAN: Deep Transfer Learning with Joint Adaptation Networks

• ICML-17 DiscoGAN: Learning to Discover Cross-Domain Relations with Generative Adversarial Networks

• JMLR-16 DANN: Domain-adversarial training of neural networks

• JMLR-16 Distribution-Matching Embedding for Visual Domain Adaptation

• ICML-15 DAN: Learning Transferable Features with Deep Adaptation Networks

• ICML-15 GRL: Unsupervised Domain Adaptation by Backpropagation

• ICML-14 DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition

AAAI && IJCAI && ICLR

• 20171210 AAAI-18 Learning to Generalize: Meta-Learning for Domain Generalization

• 201711 ICLR-18 GENERALIZING ACROSS DOMAINS VIA CROSS-GRADIENT TRAINING

• 20180116 ICLR-18 Stable Distribution Alignment using the Dual of the Adversarial Distance

• 20180110 AAAI-18 Wasserstein Distance Guided Representation Learning for Domain Adaptation

• AAAI-18 Learning to Generalize: Meta-Learning for Domain Generalization

• ICLR-18 generalizing across domains via cross-gradient training

• AAAI-18 Multi-Adversarial Domain Adaptation

• AAAI-16 Return of Frustratingly Easy Domain Adaptation

• AAAI-17 Distant Domain Transfer Learning

Others && Arxiv

• 20180111 arXiv Lifelong Learning for Sentiment Classification

• 20180111 arXiv Stable Distribution Alignment Using the Dual of the Adversarial Distance

• 20180110 arXiv Close Yet Discriminative Domain Adaptation

• 20180105 arXiv Optimal Bayesian Transfer Learning

• 20180105 arXiv Heterogeneous transfer learning

• 20171218 arXiv Partial Transfer Learning with Selective Adversarial Networks

• 20171216 arXiv Zero-Shot Deep Domain Adaptation

• 20171214 arXiv Investigating the Impact of Data Volume and Domain Similarity on Transfer Learning Applications

• 201710 Google：Learning Transferable Architectures for Scalable Image Recognition

• 201710 Domain Adaptation in Computer Vision Applications record of some domain adaptation work。  

• 201707 Adversarial Representation Learning For Domain Adaptation

• 201707 Mutual Alignment Transfer Learning

• 201708 Learning Invariant Riemannian Geometric Representations Using Deep Nets

• 20170812 HKUST：Learning To TransferCombining incremental learning and transfer learning

• CoRR abs/1711.09020 (2017) StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation

• CoRR abs/1707.01217 (2017) Wasserstein Distance Guided Representation Learning for Domain Adaptation

• 2017 Google: Learning Transferable Architectures for Scalable Image Recognition

• CoRR abs/1603.04779 (2016) AdaBN: Revisiting batch normalization for practical domain adaptation

• CoRR abs/1610.04420 (2016) Theoretical Analysis of Domain Adaptation with Optimal Transport

• KDD-15 Transitive Transfer Learning

• CoRR abs/1412.3474 (2014) Deep Domain Confusion(DDC): Maximizing for Domain Invariance

• Deep Transfer Learning with Joint Adaptation Networks， code

• Distilling the knowledge in a neural network (2015), G. Hinton et al. [pdf]

• Deep neural networks are easily fooled: High confidence predictions for unrecognizable images (2015), A. Nguyen et al. [pdf]

• How transferable are features in deep neural networks? (2014), J. Yosinski et al. [pdf]

• Visualizing and understanding convolutional networks (2014), M. Zeiler and R. Fergus [pdf]

• Decaf: A deep convolutional activation feature for generic visual recognition (2014), J. Donahue et al. [pdf]

Action Recognition

Visual event recognition in videos by learning from web data
Importance weighted least-squares probabilistic classifier for covariate shiftAdaptation with application to human activity recognition
Cost-sensitive Boosting for Concept Drift
Boosting for Transfer Learning
Interactive Event Search Through Transfer Learning
Cross-Domain Activity Recognition
Recognizing Activities in Multiple Contexts using Transfer Learning
Transfer Learning for Activity Recognition via Sensor Mapping
Cross-View Action Recognition via View Knowledge Transfer
Domain adaptation via transfer component analysis
Topology Preserving Domain Adaptation for Addressing Subject Based Variability in SEMG Signal
Remember and Transfer what you have Learned �Recognizing Composite Activities based on Activity Spotting
Transferring Knowledge of Activity Recognition across Sensor Networks
Activity Recognition Based on Home to Home Transfer Learning
Transferring Activities: Updating Human Behavior Analysis
Cross-mobile ELM based Activity Recognition
Cross-People Mobile-Phone Based Activity Recognition
Self-taught clustering
Self-taught Learning: Transfer Learning from Unlabeled Data
Transferred Dimensionality Reduction
Deep Transfer via Second-Order Markov Logic
Automatic transfer of activity recognition capabilities between body-worn motion sensors: training newcomers to recognize locomotion
Cross-Dataset Action Detection
Activity Recognition from Physiological Data using Conditional Random Fields
Heterogeneous Transfer Learning with RBMs
Hong J H, Ramos J, Dey A K. Toward Personalized Activity Recognition Systems With a Semipopulation Approach[J]. IEEE Transactions on Human-Machine Systems, 2016, 46(1): 101-112.
Morales F J O, Roggen D. Deep convolutional feature transfer across mobile activity recognition domains, sensor modalities and locations[C]//Proceedings of the 2016 ACM International Symposium on Wearable Computers. ACM, 2016: 92-99.

---

Survey

https://www.youtube.com/watch?v=qD6iD4TFsdQ

• Transitive transfer learning
• Distant Domain Transfer Learning
• A survey on transfer learning
• A survey on transfer learning
• Cross-dataset recognition: a survey
• A survey on multi-task learning
• Transfer Learning and Reinforcement Learning。
• A survey of multi-domain transfer learning
• Computer Vision domain adaptation survey。
• Theoretical Analysis

---

Codes && Blogs

Here

• 201711 A good practical tutorial for deep learning and transfer learning How to Retrain Inception's Final Layer for New Categories

---

Scholars

• Qiang Yang：IEEE/AAAI/IAPR/AAAS fellow。[Google scholar]
• Sinno Jialin Pan：[Google scholar]
• Wenyuan Dai：
• Lixin Duan
• Fuzhen Zhuang：[Google scholar]
• Mingsheng Long：[Google scholar]

---

Popular methods

• Transfer component analysis, TCA

  • Domain adaptation via tranfer component analysis

• joint distribution adaptation，JDA

  • Transfer Feature Learning with Joint Distribution Adaptation

• Geodesic flow kernel, GFK

  • Geodesic flow kernel for unsupervised domain adaptation

• Transfer Kernel Learning, TKL

  • Domain invariant transfer kernel learning

• Deep Adaptation Network, DAN

  • learning transferable features with deep adaptation networks

• Joint Adaptation Network, JAN

  • Deep Transfer Learning with Joint Adaptation Networks