JULIE THERIOT Professor of Biology, University of Washington and Chief Scientist at the Allen Institute of Cell Science) presented the imaging keynote “_Harnessing the Power of Deep Learning for Light Microscopy at a Large Scale._” The mission of the Allen Institute for Cell Science is to create dynamic and multi-scale visual models of cell organization, dynamics and activities that capture experimental observation, theory and prediction to understand and predict cellular behavior in its normal, regenerative, and pathological contexts. We want to learn what are cells doing in a moment in time so that we can learn about the past and perhaps predict the future. Learning about cell organization in relation to function can help us understand how cells transition from one state to another - in the context of differentiation, regeneration and disease. The Allen Institute is dedicated to team science and open science and is working to interact with the community. To that end we have developed the Allen Cell Explorer (allencell.org [http://allencell.org]) which is open and available to all. The Allen Cell Explorer Tool Kit contains information on gene editing in the Cell Creator, automated robotic microscopy in order to generate images in the Cell Image Generator, image analysis tools in Cell Image Analyzer and then tools for visualizing images and running simulations in the Cell Image Visualizer and Cell Image Simulator. The institute has produced over 50 hiPSC cell lines which are GFP-tagged with various cellular and organelle labels which are in use and freely available to the scientific community (20 more in development). Likewise, image data collecting methods are available online along with image analysis tools. The team at Allen Institute of Cell Science is taking advantage of the power of deep learning. In last couple years, usable software packages for deep learning in cell biology have taken off. We can now use deep convolutional neural networks for analyzing light microscopy image data to try to understand quantitatively features of cells undergoing complex behaviors. Three image analysis projects will be presented: label-free imaging (https://www.nature.com/articles/s41592-018-0111-2); improving segmentation (https://www.allencell.org/segmenter.html); and improving image resolution (new project). Our goal is to facilitate the transformation throughout the entire the cell biology research community to go from performing cell experiments and observations to high quality images to information. And subsequently to disseminate the information in a way that’s quantitative and robust and that is also completely transparent and reproducible and accessible to all the members of the research community.