Multi-Label Gold Asymmetric Loss Correction with Single-Label Regulators
This addresses the issue of costly and error-prone annotation in multi-label datasets, offering a robust solution for applications like image classification, though it is incremental as it builds on existing asymmetric loss methods.
The paper tackles the problem of noisy labels in multi-label learning by proposing GALC-SLR, a method that uses single-label samples to estimate a noise confusion matrix and apply asymmetric loss correction, resulting in up to 28.67% mean average precision improvement on MS-COCO.
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.