Revisiting Long-tailed Image Classification: Survey and Benchmarks with New Evaluation Metrics

TL;DR

This paper surveys long-tailed image classification, introduces new evaluation benchmarks with evolving distributions, and re-evaluates existing methods to better reflect real-world performance.

Contribution

It proposes novel evaluation metrics and benchmarks for long-tailed classification, and categorizes existing methods based on their training procedures.

Findings

New benchmarks with evolving distributions for testing algorithms.

Re-evaluation of methods on CIFAR datasets using proposed metrics.

Guidance for selecting data rebalancing techniques.

Abstract

Recently, long-tailed image classification harvests lots of research attention, since the data distribution is long-tailed in many real-world situations. Piles of algorithms are devised to address the data imbalance problem by biasing the training process towards less frequent classes. However, they usually evaluate the performance on a balanced testing set or multiple independent testing sets having distinct distributions with the training data. Considering the testing data may have arbitrary distributions, existing evaluation strategies are unable to reflect the actual classification performance objectively. We set up novel evaluation benchmarks based on a series of testing sets with evolving distributions. A corpus of metrics are designed for measuring the accuracy, robustness, and bounds of algorithms for learning with long-tailed distribution. Based on our benchmarks, we re-evaluate the performance of existing methods on CIFAR10 and CIFAR100 datasets, which is valuable for guiding the selection of data rebalancing techniques. We also revisit existing methods and categorize them into four types including data balancing, feature balancing, loss balancing, and prediction balancing, according the focused procedure during the training pipeline.