Abstract:  Labeling data is expensive and training a deep learning model using synthetically generated data is a promising approach. In this project, students will learn how to synthesize training for the co-segmentation task.

Contact:Weihao Li