mlsanityData to 14 Oct 2026

Learning to Detect Noisy Labels Using Model-Based Features

arXiv 2212.13767 ↗First posted 28 Dec 2022cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

Label noise is ubiquitous in various machine learning scenarios such as self-labeling with model predictions and erroneous data annotation. Many existing approaches are based on heuristics such as sample losses, which might not be flexible enough to achieve optimal solutions. Meta learning based methods address this issue by learning a data selection function, but can be hard to optimize. In light of these pros and cons, we propose Selection-Enhanced Noisy label Training (SENT) that does not rely on meta learning while having the flexibility of being data-driven. SENT transfers the noise distribution to a clean set and trains a model to distinguish noisy labels from clean ones using model-based features. Empirically, on a wide range of tasks including text classification and speech recognition, SENT improves performance over strong baselines under the settings of self-training and label corruption.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 30 of its 45 references that are in the map. First posted on arXiv 28 Dec 2022.

Builds on

30 of 45 references.

References · 30

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2015Character-level Convolutional Networks for Text Classificationcited by 2,330
  3. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,085
  4. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,311
  5. 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,059
  6. 2020Meta-Learning in Neural Networks: A Surveycited by 950
  7. 2018Learning to Reweight Examples for Robust Deep Learningcited by 675
  8. 2020Learning from Noisy Labels with Deep Neural Networks: A Surveycited by 582
  9. 2021FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labelingcited by 540
  10. 2019Confidence Regularized Self-Trainingcited by 419
  11. 2019How does Disagreement Help Generalization against Label Corruption?cited by 390
  12. 2017AISHELL-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baselinecited by 373
18 more
  1. 2017Snorkel: Rapid Training Data Creation with Weak Supervisioncited by 372
  2. 2014Training Deep Neural Networks on Noisy Labels with Bootstrappingcited by 322
  3. 2017WebVision Database: Visual Learning and Understanding from Web Datacited by 262
  4. 2019Billion-scale semi-supervised learning for image classificationcited by 232
  5. 2017CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noisecited by 181
  6. 2019SELF: Learning to Filter Noisy Labels with Self-Ensemblingcited by 164
  7. 2020Improved Noisy Student Training for Automatic Speech Recognitioncited by 141
  8. 2016Harnessing Deep Neural Networks with Logic Rulescited by 135
  9. 2019Deep Self-Learning From Noisy Labelscited by 122
  10. 2019Meta Label Correction for Noisy Label Learningcited by 108
  11. 2016Stance Detection with Bidirectional Conditional Encodingcited by 93
  12. 2019Learning to Self-Train for Semi-Supervised Few-Shot Classificationcited by 83
  13. 2020Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label Noisecited by 82
  14. 2020Learning from Rules Generalizing Labeled Exemplarscited by 59
  15. 2018Snorkel DryBell: A Case Study in Deploying Weak Supervision at Industrial Scalecited by 52
  16. 2021Self-Training with Weak Supervisioncited by 49
  17. 2020Adaptive Self-training for Few-shot Neural Sequence Labelingcited by 18
  18. 2021On sensitivity of meta-learning to support datacited by 15

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 3; 2026: 2.

6 papers.

  1. 2025QLASS: Boosting Language Agent Inference via Q-Guided Stepwise Searchcited by 20
  2. 2024SparseCL: Sparse Contrastive Learning for Contradiction Retrievalcited by 2
  3. 2025DINA: A Dual Defense Framework Against Internal Noise and External Attacks in Natural Language Processingcited by 0
  4. 2025Weed Out, Then Harvest: Dual Low-Rank Adaptation is an Effective Noisy Label Detector for Noise-Robust Learningcited by 0
  5. 2026Noise-Aware Framework for Correcting Corrupted Labelscited by 0
  6. 2026A Data-Centric Framework for Detecting and Correcting Corrupted Labelscited by 0