Research map

Label-efficient and robust learning: research map

Learning with few, noisy or unlabeled data and under distribution shift: semi- and self-supervised learning, domain adaptation, OOD detection, noisy labels.

10,657 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026), in 9 topics. Within all five venues, its share shrank from 10.9% in 2023–24 to 6.7% in 2025–26 (2,869 → 3,055 papers at ICML, NeurIPS, CVPR and AAAI, the venues with data for all four years).

Explore Label-efficient and robust learning in the interactive map

Topics

• Recognition and long-tail · 4,953 papers
  Contrastive representation theory, Correspondence and causal representation learning, Dense features and foundation priors
• Domain adaptation and generalization · 1,330 papers
  Conditional distribution alignment, Feature alignment for unsupervised adaptation, Source-free domain adaptation
• Self-supervised and contrastive · 988 papers
  Pretext-task visual pre-training, Self-supervision in supervised pipelines, Self-supervised neuroimaging representations
• Zero-/few-shot and meta-learning · 784 papers
  Generative feature synthesis for zero-shot, Zero-shot robustness and transfer, Robust fine-tuning of zero-shot models
• Semi-supervised learning · 623 papers
  Holistic semi-supervised pipelines, Self-training theory, Safe and realistic semi-supervised learning
• OOD and anomaly detection · 614 papers
  Provable and feature-based OOD detection, Spurious-aware OOD detection, Out-of-distribution generalization theory
• Knowledge distillation · 544 papers
  Rethinking distillation objectives, Distillation with causal and adaptive guidance, Few-shot and metric distillation
• Noisy labels · 462 papers
  Sample selection and co-training, Noise-aware loss and pairwise methods, Contrastive learning on noisy data
• Test-time adaptation · 359 papers
  Test-time adaptation of vision-language models, Realistic and stable test-time adaptation, Extrapolation and single-sample adaptation

Most cited papers in Label-efficient and robust learning

• The Unreasonable Effectiveness of Deep Features as a Perceptual Metric (CVPR 2018 · 13,855 citations)
• Momentum Contrast for Unsupervised Visual Representation Learning (CVPR 2020 · 12,585 citations)
• ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019 · 7,708 citations)
• A Simple Framework for Contrastive Learning of Visual Representations (ICML 2020 · 7,181 citations)
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks (ICML 2017 · 6,288 citations)
• Prototypical Networks for Few-shot Learning (NeurIPS 2017 · 5,178 citations)
• Adversarial Discriminative Domain Adaptation (CVPR 2017 · 5,087 citations)
• Unsupervised Feature Learning via Non-Parametric Instance Discrimination (CVPR 2018 · 3,635 citations)
• ChestX-ray8: Hospital-Scale Chest X-Ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases (CVPR 2017 · 3,529 citations)
• Bootstrap Your Own Latent - A New Approach to Self-Supervised Learning (NeurIPS 2020 · 3,384 citations)
• SphereFace: Deep Hypersphere Embedding for Face Recognition (CVPR 2017 · 3,010 citations)
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics (CVPR 2018 · 3,007 citations)