Video Analysis: Efficiently Tracking and Detecting Life Cycle Phase Transitions for Live Cells

To design frameworks and systems which close the gap between computer vision algorithms and human performance for analyzing live cells observed in videos.

To tackle many of the current roadblocks with live video imaging, this group published a hybrid human-machine crowdsourcing tool that provides a scalable and accurate method to detect the life cycle of cells in video imaging. This tool has both improved processing and capability to minimize the need for human expertise. They also created shared and labeled datasets for benchmarking algorithms to help improve collaboration in this space.