Towards Calibrated Hyper-Sphere Representation via Distribution Overlap Coefficient for Long-tailed Learning

TL;DR

This paper introduces a novel approach for long-tailed learning by quantifying class dominance using a distribution overlap coefficient within a hyper-sphere representation, leading to improved calibration and performance.

Contribution

It proposes the first measurement of representation quality via distribution overlap, a vMF classifier, and a zero-cost post-training calibration method for long-tailed tasks.

Findings

Achieves state-of-the-art results on long-tailed image classification.

Outperforms previous methods with significant margins.

Effective across multiple tasks like segmentation and classification.

Abstract

Long-tailed learning aims to tackle the crucial challenge that head classes dominate the training procedure under severe class imbalance in real-world scenarios. However, little attention has been given to how to quantify the dominance severity of head classes in the representation space. Motivated by this, we generalize the cosine-based classifiers to a von Mises-Fisher (vMF) mixture model, denoted as vMF classifier, which enables to quantitatively measure representation quality upon the hyper-sphere space via calculating distribution overlap coefficient. To our knowledge, this is the first work to measure representation quality of classifiers and features from the perspective of distribution overlap coefficient. On top of it, we formulate the inter-class discrepancy and class-feature consistency loss terms to alleviate the interference among the classifier weights and align features with classifier weights. Furthermore, a novel post-training calibration algorithm is devised to zero-costly boost the performance via inter-class overlap coefficients. Our method outperforms previous work with a large margin and achieves state-of-the-art performance on long-tailed image classification, semantic segmentation, and instance segmentation tasks (e.g., we achieve 55.0\% overall accuracy with ResNetXt-50 in ImageNet-LT). Our code is available at https://github.com/VipaiLab/vMF\_OP.