Joint Optimization Framework for Learning with Noisy Labels

Abstract

Deep neural networks (DNNs) trained on large-scale datasets have exhibited significant performance in image classification. Many large-scale datasets are collected from websites, however they tend to contain inaccurate labels that are termed as noisy labels. Training on such noisy labeled datasets causes performance degradation because DNNs easily overfit to noisy labels. To overcome this problem, we propose a joint optimization framework of learning DNN parameters and estimating true labels. Our framework can correct labels during training by alternating update of network parameters and labels. We conduct experiments on the noisy CIFAR-10 datasets and the Clothing1M dataset. The results indicate that our approach significantly outperforms other state-of-the-art methods.

arXiv commentTo appear at CVPR 2018 (poster), including supplementary material

References in corpus (2)

• Learning Discrete Representations via Information Maximizing Self-Augmented Training
• Learning Deep Networks from Noisy Labels with Dropout Regularization

Cited by in corpus (32)

• Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels
• Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels
• Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting
• Masking: A New Perspective of Noisy Supervision
• How does Disagreement Help Generalization against Label Corruption?
• Sample Selection with Uncertainty of Losses for Learning with Noisy Labels
• EEG-Based Emotion Recognition Using Regularized Graph Neural Networks
• IMAE for Noise-Robust Learning: Mean Absolute Error Does Not Treat Examples Equally and Gradient Magnitude's Variance Matters
• Confidence Scores Make Instance-dependent Label-noise Learning Possible
• Training Binary Neural Networks through Learning with Noisy Supervision
• Learning with Bounded Instance- and Label-dependent Label Noise
• MetaLabelNet: Learning to Generate Soft-Labels from Noisy-Labels
• Learning Soft Labels via Meta Learning
• ProSelfLC: Progressive Self Label Correction for Training Robust Deep Neural Networks
• Improving Multi-Person Pose Estimation using Label Correction
• Weak-shot Fine-grained Classification via Similarity Transfer
• Derivative Manipulation for General Example Weighting
• Boosting Co-teaching with Compression Regularization for Label Noise
• Instance Correction for Learning with Open-set Noisy Labels
• Weakly Supervised Learning with Side Information for Noisy Labeled Images
• Robust and On-the-fly Dataset Denoising for Image Classification
• The Resistance to Label Noise in K-NN and DNN Depends on its Concentration
• NLNL: Negative Learning for Noisy Labels
• Noisy Annotation Refinement for Object Detection
• Product Image Recognition with Guidance Learning and Noisy Supervision
• Distilling effective supervision for robust medical image segmentation with noisy labels
• Detecting and Tracking Communal Bird Roosts in Weather Radar Data
• Photometric Transformer Networks and Label Adjustment for Breast Density Prediction
• Multiplicative Reweighting for Robust Neural Network Optimization
• Towards Robust Pattern Recognition: A Review
• Reliable Label Bootstrapping for Semi-Supervised Learning
• Suppressing Mislabeled Data via Grouping and Self-Attention