Learning Confident Classifiers in the Presence of Label Noise
This addresses the challenge of label noise in deep learning, particularly critical for domains like medical imaging where multiple expert annotations are common but subjective, though it appears incremental as it builds on existing noise-handling methods.
The paper tackles the problem of training deep neural networks with noisy labels by proposing a probabilistic model with information-based regularization to recover ground-truth labels, showing it outperforms state-of-the-art methods on classification tasks like noisy MNIST and CIFAR-10N and segmentation tasks on medical datasets like LIDC and RIGA.
The success of Deep Neural Network (DNN) models significantly depends on the quality of provided annotations. In medical image segmentation, for example, having multiple expert annotations for each data point is common to minimize subjective annotation bias. Then, the goal of estimation is to filter out the label noise and recover the ground-truth masks, which are not explicitly given. This paper proposes a probabilistic model for noisy observations that allows us to build a confident classification and segmentation models. To accomplish it, we explicitly model label noise and introduce a new information-based regularization that pushes the network to recover the ground-truth labels. In addition, for segmentation task we adjust the loss function by prioritizing learning in high-confidence regions where all the annotators agree on labeling. We evaluate the proposed method on a series of classification tasks such as noisy versions of MNIST, CIFAR-10, Fashion-MNIST datasets as well as CIFAR-10N, which is real-world dataset with noisy human annotations. Additionally, for segmentation task, we consider several medical imaging datasets, such as, LIDC and RIGA that reflect real-world inter-variability among multiple annotators. Our experiments show that our algorithm outperforms state-of-the-art solutions for the considered classification and segmentation problems.