Speaker: Dr. Dushan Wadduwage, Old Dominion University
Abstract:
Living systems at the microscopic scale look and behave differently from typical objects we observe daily. Consequently, image data generated at these two scales also differ. Yet deep learning-based vision models -originally designed for typical computer vision applications- are commonly applied in imaging experiments in microscopy. This talk will discuss what makes these data different and how best to adopt state-of-the-art deep-learning models in microscopy experiments. Using examples from fluorescence, two-photon, and quantitative phase microscopy, we will discuss how to use machine learning approaches not only to process image data and derive biologically relevant conclusions, but also to design more efficient imaging systems.