Difficulty-aware Balancing Margin Loss for Long-tailed Recognition

TL;DR

This paper introduces a difficulty-aware balancing margin loss that addresses both class imbalance and individual sample difficulty, improving recognition accuracy in long-tailed datasets.

Contribution

It proposes a novel DBM loss that incorporates class-wise and instance-wise margins, enhancing discriminativity for hard samples in imbalanced data.

Findings

Consistently improves performance across long-tailed benchmarks.

Effectively balances class bias and instance difficulty.

Seamlessly integrates with existing methods.

Abstract

When trained with severely imbalanced data, deep neural networks often struggle to accurately recognize classes with only a few samples. Previous studies in long-tailed recognition have attempted to rebalance biased learning using known sample distributions, primarily addressing different classification difficulties at the class level. However, these approaches often overlook the instance difficulty variation within each class. In this paper, we propose a difficulty-aware balancing margin (DBM) loss, which considers both class imbalance and instance difficulty. DBM loss comprises two components: a class-wise margin to mitigate learning bias caused by imbalanced class frequencies, and an instance-wise margin assigned to hard positive samples based on their individual difficulty. DBM loss improves class discriminativity by assigning larger margins to more difficult samples. Our method seamlessly combines with existing approaches and consistently improves performance across various long-tailed recognition benchmarks.