CVPR 2020 accepted papers on Label-efficient and robust learning
219 of the 1,467 papers accepted at CVPR 2020 (14.9%) are on Label-efficient and robust learning, compared with 15.9% at CVPR 2019. They fall into 8 topics.
See CVPR 2020 in the interactive map
Submitting to CVPR? 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 · 88 papers
- Circle Loss: A Unified Perspective of Pair Similarity Optimization (956 citations)
- BBN: Bilateral-Branch Network With Cumulative Learning for Long-Tailed Visual Recognition (905 citations)
- Suppressing Uncertainties for Large-Scale Facial Expression Recognition (722 citations)
- Relation-Aware Global Attention for Person Re-Identification (655 citations)
- CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition (590 citations)
- Cross-Modality Person Re-Identification With Shared-Specific Feature Transfer (431 citations)
- Proxy Anchor Loss for Deep Metric Learning (411 citations)
- Unsupervised Person Re-Identification via Multi-Label Classification (374 citations)
- Hi-CMD: Hierarchical Cross-Modality Disentanglement for Visible-Infrared Person Re-Identification (370 citations)
- Hierarchical Clustering With Hard-Batch Triplet Loss for Person Re-Identification (319 citations)
- Unsupervised Person Re-Identification via Softened Similarity Learning (315 citations)
- Hyperbolic Image Embeddings (312 citations)
and 76 more in the interactive map.
Domain adaptation and generalization · 45 papers
- FDA: Fourier Domain Adaptation for Semantic Segmentation (1,119 citations)
- Model Adaptation: Unsupervised Domain Adaptation Without Source Data (502 citations)
- Learning to Learn Single Domain Generalization (468 citations)
- Style Normalization and Restitution for Generalizable Person Re-Identification (439 citations)
- Unsupervised Intra-Domain Adaptation for Semantic Segmentation Through Self-Supervision (387 citations)
- AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive Person Re-Identification (320 citations)
- Unsupervised Domain Adaptation via Structurally Regularized Deep Clustering (319 citations)
- Single-Side Domain Generalization for Face Anti-Spoofing (315 citations)
- Gradually Vanishing Bridge for Adversarial Domain Adaptation (281 citations)
- Learning Texture Invariant Representation for Domain Adaptation of Semantic Segmentation (278 citations)
- Differential Treatment for Stuff and Things: A Simple Unsupervised Domain Adaptation Method for Semantic Segmentation (250 citations)
- Universal Source-Free Domain Adaptation (240 citations)
and 33 more in the interactive map.
Semi-supervised learning · 21 papers
- Semi-Supervised Semantic Segmentation With Cross-Consistency Training (981 citations)
- Towards Discriminability and Diversity: Batch Nuclear-Norm Maximization Under Label Insufficient Situations (418 citations)
- A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection (167 citations)
- ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation (166 citations)
- FocalMix: Semi-Supervised Learning for 3D Medical Image Detection (153 citations)
- TransMatch: A Transfer-Learning Scheme for Semi-Supervised Few-Shot Learning (135 citations)
- Learning From Synthetic Animals (123 citations)
- Gait Recognition via Semi-supervised Disentangled Representation Learning to Identity and Covariate Features (111 citations)
- 3FabRec: Fast Few-Shot Face Alignment by Reconstruction (80 citations)
- Flow Contrastive Estimation of Energy-Based Models (60 citations)
- A Semi-Supervised Assessor of Neural Architectures (58 citations)
- Deep Active Learning for Biased Datasets via Fisher Kernel Self-Supervision (53 citations)
and 9 more in the interactive map.
Zero-/few-shot and meta-learning · 21 papers
- Few-Shot Learning via Embedding Adaptation With Set-to-Set Functions (776 citations)
- Adaptive Subspaces for Few-Shot Learning (464 citations)
- Adversarial Feature Hallucination Networks for Few-Shot Learning (320 citations)
- Fine-Grained Generalized Zero-Shot Learning via Dense Attribute-Based Attention (253 citations)
- Improved Few-Shot Visual Classification (242 citations)
- Boosting Few-Shot Learning With Adaptive Margin Loss (202 citations)
- Instance Credibility Inference for Few-Shot Learning (201 citations)
- Episode-Based Prototype Generating Network for Zero-Shot Learning (181 citations)
- Domain-Aware Visual Bias Eliminating for Generalized Zero-Shot Learning (178 citations)
- Generalized Zero-Shot Learning via Over-Complete Distribution (144 citations)
- Few-Shot Open-Set Recognition Using Meta-Learning (121 citations)
- Hyperbolic Visual Embedding Learning for Zero-Shot Recognition (120 citations)
and 9 more in the interactive map.
Self-supervised and contrastive · 20 papers
- Momentum Contrast for Unsupervised Visual Representation Learning (12,585 citations)
- Self-Supervised Learning of Pretext-Invariant Representations (1,405 citations)
- On the Uncertainty of Self-Supervised Monocular Depth Estimation (278 citations)
- Self-Supervised Monocular Trained Depth Estimation Using Self-Attention and Discrete Disparity Volume (265 citations)
- Learning by Analogy: Reliable Supervision From Transformations for Unsupervised Optical Flow Estimation (188 citations)
- Adaptive Dilated Network With Self-Correction Supervision for Counting (187 citations)
- MAST: A Memory-Augmented Self-Supervised Tracker (168 citations)
- Just Go With the Flow: Self-Supervised Scene Flow Estimation (154 citations)
- Evolving Losses for Unsupervised Video Representation Learning (141 citations)
- How Useful Is Self-Supervised Pretraining for Visual Tasks? (120 citations)
- Learning Representations by Predicting Bags of Visual Words (100 citations)
- Solving Mixed-Modal Jigsaw Puzzle for Fine-Grained Sketch-Based Image Retrieval (97 citations)
and 8 more in the interactive map.
Knowledge distillation · 11 papers
- Self-Training With Noisy Student Improves ImageNet Classification (2,328 citations)
- Revisiting Knowledge Distillation via Label Smoothing Regularization (551 citations)
- Dreaming to Distill: Data-Free Knowledge Transfer via DeepInversion (532 citations)
- Online Knowledge Distillation via Collaborative Learning (329 citations)
- Regularizing Class-Wise Predictions via Self-Knowledge Distillation (320 citations)
- Heterogeneous Knowledge Distillation Using Information Flow Modeling (148 citations)
- Inter-Region Affinity Distillation for Road Marking Segmentation (139 citations)
- Explaining Knowledge Distillation by Quantifying the Knowledge (121 citations)
- Collaborative Distillation for Ultra-Resolution Universal Style Transfer (103 citations)
- Neural Networks Are More Productive Teachers Than Human Raters: Active Mixup for Data-Efficient Knowledge Distillation From a Blackbox Model (48 citations)
- Distilling Cross-Task Knowledge via Relationship Matching (28 citations)
Noisy labels · 8 papers
- Combating Noisy Labels by Agreement: A Joint Training Method with Co-Regularization (600 citations)
- Transfer Learning From Synthetic to Real-Noise Denoising With Adaptive Instance Normalization (225 citations)
- Distilling Effective Supervision From Severe Label Noise (147 citations)
- Training Noise-Robust Deep Neural Networks via Meta-Learning (88 citations)
- Global-Local GCN: Large-Scale Label Noise Cleansing for Face Recognition (72 citations)
- Learning From Web Data With Self-Organizing Memory Module (20 citations)
- Task Agnostic Robust Learning on Corrupt Outputs by Correlation-Guided Mixture Density Networks (7 citations)
- ProAlignNet: Unsupervised Learning for Progressively Aligning Noisy Contours (5 citations)
OOD and anomaly detection · 5 papers
- Generalized ODIN: Detecting Out-of-Distribution Image Without Learning From Out-of-Distribution Data (476 citations)
- Graph Embedded Pose Clustering for Anomaly Detection (204 citations)
- Old Is Gold: Redefining the Adversarially Learned One-Class Classifier Training Paradigm (190 citations)
- Deep Residual Flow for Out of Distribution Detection (78 citations)
- Background Data Resampling for Outlier-Aware Classification (48 citations)
