LGCVMLJun 18, 2019

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

arXiv:1906.07413v22095 citations
AI Analysis

This addresses the challenge of training models on class-imbalanced data for better generalization on less frequent classes, representing an incremental improvement over existing techniques.

The paper tackles the problem of poor deep learning performance on imbalanced datasets by proposing a label-distribution-aware margin loss and a deferred re-weighting training schedule, achieving improved results on benchmarks like iNaturalist 2018.

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes. We design two novel methods to improve performance in such scenarios. First, we propose a theoretically-principled label-distribution-aware margin (LDAM) loss motivated by minimizing a margin-based generalization bound. This loss replaces the standard cross-entropy objective during training and can be applied with prior strategies for training with class-imbalance such as re-weighting or re-sampling. Second, we propose a simple, yet effective, training schedule that defers re-weighting until after the initial stage, allowing the model to learn an initial representation while avoiding some of the complications associated with re-weighting or re-sampling. We test our methods on several benchmark vision tasks including the real-world imbalanced dataset iNaturalist 2018. Our experiments show that either of these methods alone can already improve over existing techniques and their combination achieves even better performance gains.

Code Implementations7 repos

Data from Papers with Code (CC-BY-SA-4.0)

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes