Long-tailed Recognition by Learning from Latent Categories

TL;DR

This paper introduces LCReg, a novel method for long-tailed image recognition that leverages shared latent features and semantic data augmentation to improve feature diversity and recognition performance.

Contribution

The paper proposes a latent category-based approach that learns class-agnostic features and enhances diversity through semantic augmentation, outperforming previous methods.

Findings

Achieves state-of-the-art results on five datasets.

Significantly outperforms previous long-tailed recognition methods.

Demonstrates the effectiveness of shared latent features and semantic augmentation.

Abstract

In this work, we address the challenging task of long-tailed image recognition. Previous long-tailed recognition methods commonly focus on the data augmentation or re-balancing strategy of the tail classes to give more attention to tail classes during the model training. However, due to the limited training images for tail classes, the diversity of tail class images is still restricted, which results in poor feature representations. In this work, we hypothesize that common latent features among the head and tail classes can be used to give better feature representation. Motivated by this, we introduce a Latent Categories based long-tail Recognition (LCReg) method. Specifically, we propose to learn a set of class-agnostic latent features shared among the head and tail classes. Then, we implicitly enrich the training sample diversity via applying semantic data augmentation to the latent features. Extensive experiments on five long-tailed image recognition datasets demonstrate that our proposed LCReg is able to significantly outperform previous methods and achieve state-of-the-art results.