Influence-Balanced Loss for Imbalanced Visual Classification

TL;DR

This paper introduces a new loss function for imbalanced visual classification that reduces overfitting caused by dominant samples, improving performance across various imbalance learning methods.

Contribution

It proposes a novel influence-balanced loss that effectively addresses overfitting in imbalanced data without being restricted to specific models or tasks.

Findings

Outperforms state-of-the-art cost-sensitive loss methods on benchmark datasets.

Compatible with various imbalance learning techniques.

Enhances model robustness against overfitting in imbalanced scenarios.

Abstract

In this paper, we propose a balancing training method to address problems in imbalanced data learning. To this end, we derive a new loss used in the balancing training phase that alleviates the influence of samples that cause an overfitted decision boundary. The proposed loss efficiently improves the performance of any type of imbalance learning methods. In experiments on multiple benchmark data sets, we demonstrate the validity of our method and reveal that the proposed loss outperforms the state-of-the-art cost-sensitive loss methods. Furthermore, since our loss is not restricted to a specific task, model, or training method, it can be easily used in combination with other recent re-sampling, meta-learning, and cost-sensitive learning methods for class-imbalance problems.