Adaptive Pseudo Label Selection for Individual Unlabeled Data by Positive and Unlabeled Learning
This addresses the challenge of pseudo-label selection in medical image segmentation, but it appears incremental as it builds on existing PU learning techniques.
The paper tackles the problem of selecting effective pseudo-labels for medical image segmentation by proposing a method that uses Positive and Unlabeled Learning to discriminate foreground and background regions on individual images, resulting in improved performance as shown in experimental results.
This paper proposes a novel pseudo-labeling method for medical image segmentation that can perform learning on ``individual images'' to select effective pseudo-labels. We introduce Positive and Unlabeled Learning (PU learning), which uses only positive and unlabeled data for binary classification problems, to obtain the appropriate metric for discriminating foreground and background regions on each unlabeled image. Our PU learning makes us easy to select pseudo-labels for various background regions. The experimental results show the effectiveness of our method.