Solving The Long-Tailed Problem via Intra- and Inter-Category Balance

TL;DR

This paper introduces a novel method that addresses the long-tailed distribution challenge in visual recognition by balancing hard examples within categories and across categories, improving performance over existing approaches.

Contribution

It proposes a gradient harmonized mechanism with category-wise adaptive precision to decouple difficulty and sample size imbalance, employing intra- and inter-category balance strategies.

Findings

Consistently outperforms existing methods on multiple datasets.

Effectively handles hard examples in tail classes and decision boundary shifts.

Improves overall accuracy in long-tailed visual recognition tasks.

Abstract

Benchmark datasets for visual recognition assume that data is uniformly distributed, while real-world datasets obey long-tailed distribution. Current approaches handle the long-tailed problem to transform the long-tailed dataset to uniform distribution by re-sampling or re-weighting strategies. These approaches emphasize the tail classes but ignore the hard examples in head classes, which result in performance degradation. In this paper, we propose a novel gradient harmonized mechanism with category-wise adaptive precision to decouple the difficulty and sample size imbalance in the long-tailed problem, which are correspondingly solved via intra- and inter-category balance strategies. Specifically, intra-category balance focuses on the hard examples in each category to optimize the decision boundary, while inter-category balance aims to correct the shift of decision boundary by taking each category as a unit. Extensive experiments demonstrate that the proposed method consistently outperforms other approaches on all the datasets.