Unlabeled Image Anaysis
Abstract
Abstract
Researcher(s): Jinyeong Chae
Many deep learning approaches have been studied for image classification in computer vision. However, in medical fields, there are not enough data to generate accurate models, and many datasets are not annotated. In this study, we present a new method that can use both unlabeled data and labeled data.
Comparison with baseline model prediction and ours
Unlabeled image analysis framework