mlsanityData to 14 Oct 2026

PENCIL: Deep Learning with Noisy Labels

arXiv 2202.08436 ↗First posted 17 Feb 2022cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

Deep learning has achieved excellent performance in various computer vision tasks, but requires a lot of training examples with clean labels. It is easy to collect a dataset with noisy labels, but such noise makes networks overfit seriously and accuracies drop dramatically. To address this problem, we propose an end-to-end framework called PENCIL, which can update both network parameters and label estimations as label distributions. PENCIL is independent of the backbone network structure and does not need an auxiliary clean dataset or prior information about noise, thus it is more general and robust than existing methods and is easy to apply. PENCIL can even be used repeatedly to obtain better performance. PENCIL outperforms previous state-of-the-art methods by large margins on both synthetic and real-world datasets with different noise types and noise rates. And PENCIL is also effective in multi-label classification tasks through adding a simple attention structure on backbone networks. Experiments show that PENCIL is robust on clean datasets, too.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 22 of its 36 references that are in the map. First posted on arXiv 17 Feb 2022.

Builds on

22 of 36 references.

References · 22

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,442
  4. 2016Identity Mappings in Deep Residual Networkscited by 1,965
  5. 2016Understanding deep learning requires rethinking generalizationcited by 1,674
  6. 2018Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labelscited by 997
  7. 2017Revisiting Unreasonable Effectiveness of Data in Deep Learning Eracited by 919
  8. 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 665
  9. 2017Robust Loss Functions under Label Noise for Deep Neural Networkscited by 443
  10. 2014Training Deep Neural Networks on Noisy Labels with Bootstrappingcited by 322
  11. 2018Joint Optimization Framework for Learning with Noisy Labelscited by 293
  12. 2019Probabilistic End-to-end Noise Correction for Learning with Noisy Labelscited by 210
10 more
  1. 2017CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noisecited by 181
  2. 2018Dimensionality-Driven Learning with Noisy Labelscited by 173
  3. 2017Deep Learning is Robust to Massive Label Noisecited by 165
  4. 2018Iterative Learning with Open-set Noisy Labelscited by 108
  5. 2016Deep Label Distribution Learning with Label Ambiguitycited by 104
  6. 2017Toward Robustness against Label Noise in Training Deep Discriminative Neural Networkscited by 78
  7. 2015The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognitioncited by 48
  8. 2017Deep Learning from Noisy Image Labels with Quality Embeddingcited by 46
  9. 2019Repetitive Reprediction Deep Decipher for Semi-Supervised Learningcited by 11
  10. 2015Learning Concept Embeddings with Combined Human-Machine Expertisecited by 6

Built on by

Cited by year

Citing papers per year: 2024: 1.

1 paper.

  1. 2024A Review of Pseudo-Labeling for Computer Visioncited by 16