My research centers on demarcating objects in images (segmentation) because the task is critical to many downstream applications, such as object tracking, image registration, image classification, and image retrieval. I will begin the talk by presenting a detailed analysis of the relative strengths and weaknesses of three different image segmentation approaches: by experts, by crowd workers, and by algorithms. Then, I will describe three hybrid system designs that mix algorithm and crowd efforts. My results show that hybrid system designs can yield more accurate results than relying on algorithms or crowd workers alone and achieve expert-quality segmentations. An emphasis of my work is on biomedical/medical images in order to contribute to research which targets societyís health problems and basic science. My datasets and code are freely-available online (http://www.cs.bu.edu/~betke/BiomedicalImageSegmentation/).
BIO: Danna Gurari is a Postdoctoral Fellow at University of Texas at Austin under the supervision of Dr. Kristen Grauman. She completed her PhD at Boston University in the Image and Video Computing Group under the supervision of Dr. Margrit Betke. Dannaís research interests span computer vision, crowdsourcing, and medical/biomedical image analysis. Her current research focus is on designing prediction systems which recommend how to utilize algorithms, crowd workers, and experts for image segmentation. In 2007-2010, she worked at Boulder Imaging building custom, high performance, multi-camera recording and analysis systems for military, industrial, and academic applications. From 2005-2007, she worked at Raytheon developing software for satellite systems. Danna earned her MS in Computer Science and BS in Biomedical Engineering from Washington University in St. Louis in 2005, with her thesis on ultrasound imaging. She was awarded the 2015 Researcher Excellence Award from the Boston University computer science department, 2014 Best Paper Award for Innovative Idea at MICCAI IMIC, and 2013 Best Paper Award at WACV.