Uncertainty-based method for improving poorly labeled segmentation datasets
This addresses the challenge of data annotation quality in medical imaging or other domains where clean labels are scarce, though it is incremental as it builds on existing uncertainty methods.
The paper tackles the problem of training deep learning models for image segmentation with noisy or poorly labeled datasets by proposing a framework that identifies and corrects erroneous labels using aleatoric uncertainty estimates, resulting in improved segmentation accuracy.
The success of modern deep learning algorithms for image segmentation heavily depends on the availability of large datasets with clean pixel-level annotations (masks), where the objects of interest are accurately delineated. Lack of time and expertise during data annotation leads to incorrect boundaries and label noise. It is known that deep convolutional neural networks (DCNNs) can memorize even completely random labels, resulting in poor accuracy. We propose a framework to train binary segmentation DCNNs using sets of unreliable pixel-level annotations. Erroneously labeled pixels are identified based on the estimated aleatoric uncertainty of the segmentation and are relabeled to the true value.