Abstract:Existing research on learning with noisy labels mainly focuses on synthetic label noise. Synthetic noise, though has clean structures which greatly enabled statistical analyses, often fails to model real-world noise patterns. The recent literature has observed several efforts to offer real-world noisy datasets, yet the existing efforts suffer from two caveats: (1) The lack of ground-truth verification makes it hard to theoretically study the property and treatment of real-world label noise; (2) These efforts are often of large scales, which may result in unfair comparisons of robust methods within reasonable and accessible computation power. To better understand real-world label noise, it is crucial to build controllable and moderate-sized real-world noisy datasets with both ground-truth and noisy labels. This work presents two new benchmark datasets CIFAR-10N, CIFAR-100N, equipping the training datasets of CIFAR-10, CIFAR-100 with human-annotated real-world noisy labels we collected from Amazon Mechanical Turk. We quantitatively and qualitatively show that real-world noisy labels follow an instance-dependent pattern rather than the classically assumed and adopted ones (e.g., class-dependent label noise). We then initiate an effort to benchmarking a subset of the existing solutions using CIFAR-10N and CIFAR-100N. We further proceed to study the memorization of correct and wrong predictions, which further illustrates the difference between human noise and class-dependent synthetic noise. We show indeed the real-world noise patterns impose new and outstanding challenges as compared to synthetic label noise. These observations require us to rethink the treatment of noisy labels, and we hope the availability of these two datasets would facilitate the development and evaluation of future learning with noisy label solutions. Datasets and leaderboards are available at <a class="link-external link-http" href="http://noisylabels.com" rel="external noopener nofollow">this http URL</a>.

Review–A Survey of Learning from Noisy Labels

A Survey of Label-Noise Deep Learning for Medical Image Analysis

Rethinking Noisy Label Learning in Real-world Annotation Scenarios from the Noise-type Perspective

Noisy Label Processing for Classification: A Survey

A Survey of Label-noise Representation Learning: Past, Present and Future

Learning with Noisy Labels Via Self-supervised Adversarial Noisy Masking

Deep Active Learning in the Presence of Label Noise: A Survey

Deep learning with noisy labels: Exploring techniques and remedies in medical image analysis

An Empirical Study on Noisy Label Learning for Program Understanding

Learning from Noisy Labels with Decoupled Meta Label Purifier

Towards Understanding Deep Learning from Noisy Labels with Small-Loss Criterion

Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey

Deep learning with noisy labels in medical prediction problems: a scoping review

Noisy Label Learning for Large-scale Medical Image Classification

Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotations

Learning to Detect Noisy Labels Using Model-Based Features

A Convergence Path to Deep Learning on Noisy Labels

On Better Detecting and Leveraging Noisy Samples for Learning with Severe Label Noise

Learning from Noisy Labels Via Meta Credible Label Elicitation.

Error-Bounded Correction of Noisy Labels

Mitigating the impact of mislabeled data on deep predictive models: an empirical study of learning with noise approaches in software engineering tasks