NeurIPS 2018 accepted papers on Label-efficient and robust learning
73 of the 1,008 papers accepted at NeurIPS 2018 (7.2%) are on Label-efficient and robust learning, compared with 6.7% at NeurIPS 2017. They fall into 8 topics.
See NeurIPS 2018 in the interactive map
Submitting to NeurIPS? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Recognition and long-tail · 32 papers
- Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (431 citations)
- Towards Robust Interpretability with Self-Explaining Neural Networks (414 citations)
- MiME: Multilevel Medical Embedding of Electronic Health Records for Predictive Healthcare (178 citations)
- Neural Voice Cloning with a Few Samples (178 citations)
- Explanations based on the Missing: Towards Contrastive Explanations with Pertinent Negatives (148 citations)
- Batch-Instance Normalization for Adaptively Style-Invariant Neural Networks (125 citations)
- Model Agnostic Supervised Local Explanations (124 citations)
- FRAGE: Frequency-Agnostic Word Representation (110 citations)
- Reducing Network Agnostophobia (79 citations)
- DeepPINK: reproducible feature selection in deep neural networks (66 citations)
- CapProNet: Deep Feature Learning via Orthogonal Projections onto Capsule Subspaces (56 citations)
- Maximum-Entropy Fine Grained Classification (49 citations)
and 20 more in the interactive map.
Domain adaptation and generalization · 13 papers
- Conditional Adversarial Domain Adaptation (1,425 citations)
- MetaReg: Towards Domain Generalization using Meta-Regularization (391 citations)
- Generalizing to Unseen Domains via Adversarial Data Augmentation (361 citations)
- Adversarial Multiple Source Domain Adaptation (325 citations)
- Co-regularized Alignment for Unsupervised Domain Adaptation (115 citations)
- One-Shot Unsupervised Cross Domain Translation (104 citations)
- A Unified Feature Disentangler for Multi-Domain Image Translation and Manipulation (102 citations)
- Extracting Relationships by Multi-Domain Matching (62 citations)
- Adapted Deep Embeddings: A Synthesis of Methods for k-Shot Inductive Transfer Learning (48 citations)
- Unsupervised Adversarial Invariance (22 citations)
- The Global Anchor Method for Quantifying Linguistic Shifts and Domain Adaptation (20 citations)
- Unsupervised Image-to-Image Translation Using Domain-Specific Variational Information Bound (20 citations)
and 1 more in the interactive map.
Noisy labels · 9 papers
- Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels (1,467 citations)
- Co-teaching: Robust training of deep neural networks with extremely noisy labels (1,182 citations)
- Masking: A New Perspective of Noisy Supervision (170 citations)
- Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise (114 citations)
- Modern Neural Networks Generalize on Small Data Sets (111 citations)
- Collaborative Learning for Deep Neural Networks (100 citations)
- Robustness of conditional GANs to noisy labels (65 citations)
- Robot Learning in Homes: Improving Generalization and Reducing Dataset Bias (60 citations)
- Removing the Feature Correlation Effect of Multiplicative Noise (3 citations)
Zero-/few-shot and meta-learning · 8 papers
- Delta-encoder: an effective sample synthesis method for few-shot object recognition (289 citations)
- MetaGAN: An Adversarial Approach to Few-Shot Learning (277 citations)
- TADAM: Task dependent adaptive metric for improved few-shot learning (194 citations)
- Generalized Zero-Shot Learning with Deep Calibration Network (168 citations)
- Learning Deep Disentangled Embeddings With the F-Statistic Loss (116 citations)
- Domain-Invariant Projection Learning for Zero-Shot Recognition (40 citations)
- Supervising Unsupervised Learning (24 citations)
- Zero-Shot Transfer with Deictic Object-Oriented Representation in Reinforcement Learning (1 citation)
Semi-supervised learning · 5 papers
- Realistic Evaluation of Deep Semi-Supervised Learning Algorithms (205 citations)
- Semi-supervised Deep Kernel Learning: Regression with Unlabeled Data by Minimizing Predictive Variance (44 citations)
- Dual Swap Disentangling (14 citations)
- Semi-Supervised Learning with Declaratively Specified Entropy Constraints (4 citations)
- The Pessimistic Limits and Possibilities of Margin-based Losses in Semi-supervised Learning (4 citations)
Knowledge distillation · 4 papers
- Knowledge Distillation by On-the-Fly Native Ensemble (366 citations)
- KDGAN: Knowledge Distillation with Generative Adversarial Networks (132 citations)
- Paraphrasing Complex Network: Network Compression via Factor Transfer (86 citations)
- Learning to Teach with Dynamic Loss Functions (83 citations)
