Learning From Noisy Large-Scale Datasets With Minimal Supervision

Abstract

We present an approach to effectively use millions of images with noisy annotations in conjunction with a small subset of cleanly-annotated images to learn powerful image representations. One common approach to combine clean and noisy data is to first pre-train a network using the large noisy dataset and then fine-tune with the clean dataset. We show this approach does not fully leverage the information contained in the clean set. Thus, we demonstrate how to use the clean annotations to reduce the noise in the large dataset before fine-tuning the network using both the clean set and the full set with reduced noise. The approach comprises a multi-task network that jointly learns to clean noisy annotations and to accurately classify images. We evaluate our approach on the recently released Open Images dataset, containing ~9 million images, multiple annotations per image and over 6000 unique classes. For the small clean set of annotations we use a quarter of the validation set with ~40k images. Our results demonstrate that the proposed approach clearly outperforms direct fine-tuning across all major categories of classes in the Open Image dataset. Further, our approach is particularly effective for a large number of classes with wide range of noise in annotations (20-80% false positive annotations).

CVPR 2017

References in corpus (2)

• Distilling the Knowledge in a Neural Network
• Training Convolutional Networks with Noisy Labels

Cited by in corpus (30)

• 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
• Dimensionality-Driven Learning with Noisy Labels
• Masking: A New Perspective of Noisy Supervision
• How does Disagreement Help Generalization against Label Corruption?
• CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noise
• Iterative Learning with Open-set Noisy Labels
• Avoiding Your Teacher's Mistakes: Training Neural Networks with Controlled Weak Supervision
• CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images
• Learning Category Correlations for Multi-label Image Recognition with Graph Networks
• Deep Self-Learning From Noisy Labels
• Simple and Effective Regularization Methods for Training on Noisily Labeled Data with Generalization Guarantee
• ProSelfLC: Progressive Self Label Correction for Training Robust Deep Neural Networks
• Enhancing Robustness of On-line Learning Models on Highly Noisy Data
• A Semi-Supervised Two-Stage Approach to Learning from Noisy Labels
• Weak-shot Fine-grained Classification via Similarity Transfer
• Derivative Manipulation for General Example Weighting
• Learning with Biased Complementary Labels
• Tag Prediction at Flickr: a View from the Darkroom
• Human-Expert-Level Brain Tumor Detection Using Deep Learning with Data Distillation and Augmentation
• The Resistance to Label Noise in K-NN and DNN Depends on its Concentration
• Weakly Supervised Learning with Side Information for Noisy Labeled Images
• Deep k-NN for Noisy Labels
• Limited Gradient Descent: Learning With Noisy Labels
• Weak Novel Categories without Tears: A Survey on Weak-Shot Learning
• Mixing between the Cross Entropy and the Expectation Loss Terms
• Labeled Data Generation with Inexact Supervision
• Task-Driven Data Verification via Gradient Descent
• Semantic Segmentation with Scarce Data