Rethinking the Value of Labels for Improving Class-Imbalanced Learning

Abstract

Real-world data often exhibits long-tailed distributions with heavy class imbalance, posing great challenges for deep recognition models. We identify a persisting dilemma on the value of labels in the context of imbalanced learning: on the one hand, supervision from labels typically leads to better results than its unsupervised counterparts; on the other hand, heavily imbalanced data naturally incurs "label bias" in the classifier, where the decision boundary can be drastically altered by the majority classes. In this work, we systematically investigate these two facets of labels. We demonstrate, theoretically and empirically, that class-imbalanced learning can significantly benefit in both semi-supervised and self-supervised manners. Specifically, we confirm that (1) positively, imbalanced labels are valuable: given more unlabeled data, the original labels can be leveraged with the extra data to reduce label bias in a semi-supervised manner, which greatly improves the final classifier; (2) negatively however, we argue that imbalanced labels are not useful always: classifiers that are first pre-trained in a self-supervised manner consistently outperform their corresponding baselines. Extensive experiments on large-scale imbalanced datasets verify our theoretically grounded strategies, showing superior performance over previous state-of-the-arts. Our intriguing findings highlight the need to rethink the usage of imbalanced labels in realistic long-tailed tasks. Code is available at https://github.com/YyzHarry/imbalanced-semi-self.

NeurIPS 2020

References in corpus (15)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• A systematic study of the class imbalance problem in convolutional neural networks
• Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
• Temporal Ensembling for Semi-Supervised Learning
• Momentum Contrast for Unsupervised Visual Representation Learning
• Interpolation Consistency Training for Semi-Supervised Learning
• Data-Efficient Image Recognition with Contrastive Predictive Coding
• Adversarially Learned Inference
• Contrastive Multiview Coding
• Learning with Pseudo-Ensembles
• Decoupling Representation and Classifier for Long-Tailed Recognition
• BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition
• Selfie: Self-supervised Pretraining for Image Embedding
• Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition from a Domain Adaptation Perspective
• Manifold regularization with GANs for semi-supervised learning

Cited by in corpus (21)

• A Survey on Long-Tailed Visual Recognition
• A modular U-Net for automated segmentation of X-ray tomography images in composite materials
• Learning Imbalanced Datasets with Maximum Margin Loss
• Cross-hospital Sepsis Early Detection via Semi-supervised Optimal Transport with Self-paced Ensemble
• Rethinking Re-Sampling in Imbalanced Semi-Supervised Learning
• Self-Damaging Contrastive Learning
• Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning
• Self-supervised Mean Teacher for Semi-supervised Chest X-ray Classification
• MSE Loss with Outlying Label for Imbalanced Classification
• Robust Attentive Deep Neural Network for Exposing GAN-generated Faces
• Label-Aware Distribution Calibration for Long-tailed Classification
• Towards Resolving the Challenge of Long-tail Distribution in UAV Images for Object Detection
• Semi-supervised Long-tailed Recognition using Alternate Sampling
• Damage GAN: A Generative Model for Imbalanced Data
• ACE: Adaptive Confusion Energy for Natural World Data Distribution
• Addressing Class Variable Imbalance in Federated Semi-supervised Learning
• Subtask-dominated Transfer Learning for Long-tail Person Search
• Balanced Representation Learning for Long-tailed Skeleton-based Action Recognition
• A Semi-Supervised Framework for Misinformation Detection
• BiSTF: Bilateral-Branch Self-Training Framework for Semi-Supervised Large-scale Fine-Grained Recognition
• Self Supervision to Distillation for Long-Tailed Visual Recognition