Multi-label learning is one of the hot problems in the field of machine learning. The deep neural networks used to solve it could be quite complex and have a huge capacity. This enormous capacity, however, could also be a negative, as they tend to eventually overfit the undesirable features of the data. One such feature presented in the real-world datasets is imperfect labels. A particularly common type of label imperfection is called weak labels. This corruption is characterized not only by the presence of all relevant labels but also by the addition of some irrelevant ones. In this paper, a novel method, Co-ASL, is introduced to deal with the label noise in multi-label datasets. It combines the state-of-the-art approach for multi-label learning, ASL, with the famous robust training strategy, Co-teaching. The performance of the method is then evaluated on noisy versions of MS-COCO to show the lack of overfitting and the performance improvement over the non-robust multi-label ASL.

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.

Multi-label learning is becoming more and moreimportant as real-world data often contains multi-ple labels. The dataset used for learning such aclassifier is of great importance. Acquiring a cor-rectly labelled dataset is however a difficult task.Active learning is a method which can, given anoisy dataset, identify important instances for anexpert to label. This greatly reduces the amountof instances needed to train an accurate classi-fier, and thus reduces the cost of cleaning a noisydataset. Therefore, this paper aims to present an ac-tive learning algorithm, focused on wrongly labeleddata, combined with a deep neural network formulti-label image classification. The proposed ac-tive learning solution is divided into two measures;a mislabelling likelihood and an informativenessmeasure together with an option to identify anduse highly probable clean instances in the dataset.Experiments performed on the real world dataset,called Microsoft COCO, with 20, 40 and 60% in-jected label noise show that Multi-AL outperformsthe current state-of-the-art multi-label learning al-gorithm called ASL by 28% while only using 600labelled instances in total and 250 extracted ’clean’instances. Multi-AL additionally outperforms ran-dom sampling by 3% on average for 20 and 40%random label noise when sampling from a wronglylabelled dataset of 23k instances.