mlsanityData to 7 Oct 2026

SURE: SUrvey REcipes for building reliable and robust deep networks

arXiv 2403.00543 ↗First posted 1 Mar 2024cs.CVcited by 8

Not on the map. Why?

Abstract (arXiv)

In this paper, we revisit techniques for uncertainty estimation within deep neural networks and consolidate a suite of techniques to enhance their reliability. Our investigation reveals that an integrated application of diverse techniques--spanning model regularization, classifier and optimization--substantially improves the accuracy of uncertainty predictions in image classification tasks. The synergistic effect of these techniques culminates in our novel SURE approach. We rigorously evaluate SURE against the benchmark of failure prediction, a critical testbed for uncertainty estimation efficacy. Our results showcase a consistently better performance than models that individually deploy each technique, across various datasets and model architectures. When applied to real-world challenges, such as data corruption, label noise, and long-tailed class distribution, SURE exhibits remarkable robustness, delivering results that are superior or on par with current state-of-the-art specialized methods. Particularly on Animal-10N and Food-101N for learning with noisy labels, SURE achieves state-of-the-art performance without any task-specific adjustments. This work not only sets a new benchmark for robust uncertainty estimation but also paves the way for its application in diverse, real-world scenarios where reliability is paramount. Our code is available at urlhttps://yutingli0606.github.io/SURE/.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 69 of its 84 references that are in the map. First posted on arXiv 1 Mar 2024.

Builds on

69 of 84 references.

References · 69

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2016Densely Connected Convolutional Networkscited by 6,691
  4. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  5. 2017On Calibration of Modern Neural Networkscited by 4,860
  6. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  7. 2016Wide Residual Networkscited by 3,057
  8. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  9. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,362
  10. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
  11. 2020Sharpness-Aware Minimization for Efficiently Improving Generalizationcited by 1,404
  12. 2019MixMatch: A Holistic Approach to Semi-Supervised Learningcited by 1,390
38 more
  1. 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,325
  2. 2020Energy-based Out-of-distribution Detectioncited by 1,296
  3. 2019Class-Balanced Loss Based on Effective Number of Samplescited by 1,239
  4. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,194
  5. 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,053
  6. 2019Learning Imbalanced Datasets with Label-Distribution-Aware Margin Losscited by 987
  7. 2018Deep Anomaly Detection with Outlier Exposurecited by 939
  8. 2019Decoupling Representation and Classifier for Long-Tailed Recognitioncited by 806
  9. 2017Selective Classification for Deep Neural Networkscited by 799
  10. 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 691
  11. 2017MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labelscited by 652
  12. 2020Early-Learning Regularization Prevents Memorization of Noisy Labelscited by 414
  13. 2019BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognitioncited by 403
  14. 2019How does Disagreement Help Generalization against Label Corruption?cited by 390
  15. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  16. 2018Joint Optimization Framework for Learning with Noisy Labelscited by 293
  17. 2020Combating noisy labels by agreement: A joint training method with co-regularizationcited by 289
  18. 2020Identifying Mislabeled Data using the Area Under the Margin Rankingcited by 259
  19. 2020Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effectcited by 256
  20. 2017Decoupling "when to update" from "how to update"cited by 230
  21. 2020Rethinking the Value of Labels for Improving Class-Imbalanced Learningcited by 216
  22. 2019Probabilistic End-to-end Noise Correction for Learning with Noisy Labelscited by 210
  23. 2021Deep Deterministic Uncertainty: A Simple Baselinecited by 200
  24. 2017CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noisecited by 181
  25. 2018Dimensionality-Driven Learning with Noisy Labelscited by 173
  26. 2017Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samplescited by 147
  27. 2018Exploring Uncertainty Measures in Deep Networks for Multiple Sclerosis Lesion Detection and Segmentationcited by 146
  28. 2022Balanced Contrastive Learning for Long-Tailed Visual Recognitioncited by 143
  29. 2021Contrastive Learning based Hybrid Networks for Long-Tailed Image Classificationcited by 130
  30. 2019Deep Self-Learning From Noisy Labelscited by 122
  31. 2019A General Framework for Uncertainty Estimation in Deep Learningcited by 119
  32. 2020Confidence-Aware Learning for Deep Neural Networkscited by 113
  33. 2022Long-Tailed Recognition via Weight Balancingcited by 107
  34. 2018Bias-Reduced Uncertainty Estimation for Deep Neural Classifierscited by 106
  35. 2019Gaussian YOLOv3: An Accurate and Fast Object Detector Using Localization Uncertainty for Autonomous Drivingcited by 105
  36. 2021Learning with Feature-Dependent Label Noise: A Progressive Approachcited by 102
  37. 2021MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognitioncited by 99
  38. 2021ResLT: Residual Learning for Long-tailed Recognitioncited by 96

50 most cited of 69 All 69 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 5; 2026: 2.

8 papers.

  1. 2025Introducing Fractional Classification Loss for Robust Learning with Noisy Labelscited by 2
  2. 2025Dual-granularity Sinkhorn Distillation for Enhanced Learning from Long-tailed Noisy Datacited by 1
  3. 2026Adaptive Confidence Regularization for Multimodal Failure Detectioncited by 1
  4. 2024SAU: A Dual-Branch Network to Enhance Long-Tailed Recognition via Generative Modelscited by 0
  5. 2025TrustLoRA: Low-Rank Adaptation for Failure Detection under Out-of-distribution Datacited by 0
  6. 2025Intrinsic Dimensionality as a Model-Free Measure of Class Imbalancecited by 0
  7. 2025Learning to Clean: Reinforcement Learning for Noisy Label Correctioncited by 0
  8. 2026From Misclassifications to Outliers: Joint Reliability Assessment in Classificationcited by 0