Machine Learning Advances for 3D Image Processing and Visualization
Lauren Christopher, Ph.D.
Friday, October 20 at 12:30 p.m. in IT 252
2D image analysis has had the benefit of the recent advances in Deep Learning (DL), and this has the quickly advanced machine vision for the detection and classification of objects in a 2D scene. Three dimensional (3D) processing and time series volume data is an emerging area of DL research that can enhance the accuracy of the learning and classification. This talk will discuss the results of our recent research in transportation safety: mapping pedestrians in 3D space from a car’s camera, using machine learning for body pose estimation, and classifying the car-pedestrian interaction as dangerous or not with machine learning. A discussion of how this may be useful for medical 4D processing will be introduced.
About Lauren Christopher
Dr. Lauren Christopher attended Massachusetts Institute of Technology, where she received her S. B. and S. M. in Electrical Engineering and Computer Science in 1982, specializing in digital signal processing and chip design. She worked at RCA’s David Sarnoff Research Labs in the 1980’s developing chips for early digital television, and in the 1990s joined Thomson Consumer Electronics, where she led the first DIRECTV receiver design. She received her PhD from Purdue University in 2003, developing image segmentation techniques in 3D medical ultrasound images toward automatic cancer detection. In 2010, Dr. Christopher has been inducted into the Consumer Electronics Hall of Fame for her work in leading the development of the DIRECTV set top box. Dr. Christopher joined IUPUI in 2008, and is currently heading the Machine Intelligence Computer Vision in 3D (MICV3D) Lab in the Electrical and Computer Engineering Department. Her research continues the 3D image processing from her medical PhD work. Her research interests include real-time 3D systems, 3D sensors and 3D machine vision algorithms.