Multi-Label Gold Asymmetric Loss Correction with Single-Label Regulators

TL;DR

This paper introduces GALC-SLR, a robust multi-label learning method that corrects noisy labels using single-label regulators, significantly improving prediction accuracy on corrupted datasets.

Contribution

The paper proposes a novel noise correction technique for multi-label learning that estimates confusion matrices from single-label samples to enhance robustness against label noise.

Findings

Outperforms state-of-the-art methods under various noise levels.

Achieves up to 28.67% improvement in mean average precision on MS-COCO.

Demonstrates better generalization to unseen data.

Abstract

Multi-label learning is an emerging extension of the multi-class classification where an image contains multiple labels. Not only acquiring a clean and fully labeled dataset in multi-label learning is extremely expensive, but also many of the actual labels are corrupted or missing due to the automated or non-expert annotation techniques. Noisy label data decrease the prediction performance drastically. In this paper, we propose a novel Gold Asymmetric Loss Correction with Single-Label Regulators (GALC-SLR) that operates robust against noisy labels. GALC-SLR estimates the noise confusion matrix using single-label samples, then constructs an asymmetric loss correction via estimated confusion matrix to avoid overfitting to the noisy labels. Empirical results show that our method outperforms the state-of-the-art original asymmetric loss multi-label classifier under all corruption levels, showing mean average precision improvement up to 28.67% on a real world dataset of MS-COCO, yielding a better generalization of the unseen data and increased prediction performance.