Video analytics startup Vidrovr has been awarded a $225,000 National Science Foundation Small Business Innovation Research (SBIR) grant for research and development of an inference engine that can leverage the data and structure within video to automatically discover and train classifiers for better tagging and search ability. The startup was created by Columbia PhD candidates Joseph Ellis MS ’14 (Electrical Engineering) and Daniel Morozoff AM ’14 (GSAS) and was recently accepted to the Techstars startup accelerator. The Vidrovr team developed a news and social media video processing system called News Rover that applies machine learning algorithms to identify people and objects on screen in real time, add tags, and link text to other text- and image-based information online.