Class-Balanced Loss Based on Effective Number of Samples

Abstract

With the rapid increase of large-scale, real-world datasets, it becomes critical to address the problem of long-tailed data distribution (i.e., a few classes account for most of the data, while most classes are under-represented). Existing solutions typically adopt class re-balancing strategies such as re-sampling and re-weighting based on the number of observations for each class. In this work, we argue that as the number of samples increases, the additional benefit of a newly added data point will diminish. We introduce a novel theoretical framework to measure data overlap by associating with each sample a small neighboring region rather than a single point. The effective number of samples is defined as the volume of samples and can be calculated by a simple formula $(1-{\beta }^n)/(1-\beta )$, where $n$ is the number of samples and $\beta \in [0,1)$ is a hyperparameter. We design a re-weighting scheme that uses the effective number of samples for each class to re-balance the loss, thereby yielding a class-balanced loss. Comprehensive experiments are conducted on artificially induced long-tailed CIFAR datasets and large-scale datasets including ImageNet and iNaturalist. Our results show that when trained with the proposed class-balanced loss, the network is able to achieve significant performance gains on long-tailed datasets.

Code is available at: https://github.com/richardaecn/class-balanced-loss

References in corpus (3)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• The Devil is in the Tails: Fine-grained Classification in the Wild
• Deep Active Learning over the Long Tail

Cited by in corpus (17)

• Data Augmentation for Object Detection via Progressive and Selective Instance-Switching
• The Herbarium Challenge 2019 Dataset
• MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition
• Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning
• End-2-End COVID-19 Detection from Breath & Cough Audio
• Teacher's pet: understanding and mitigating biases in distillation
• Large-Scale Object Detection in the Wild from Imbalanced Multi-Labels
• SelectNet: Learning to Sample from the Wild for Imbalanced Data Training
• DotFAN: A Domain-transferred Face Augmentation Network for Pose and Illumination Invariant Face Recognition
• Energy Aligning for Biased Models
• Improvement of Batch Normalization in Imbalanced Data
• Learning with Multiclass AUC: Theory and Algorithms
• Metadata Shaping: Natural Language Annotations for the Tail
• Training Over-parameterized Models with Non-decomposable Objectives
• Collaging Class-specific GANs for Semantic Image Synthesis
• Class Balancing GAN with a Classifier in the Loop
• On-target Adaptation