Meeting Information
All guest lectures will be held in the Lubrano conference room on the 4th floor of the CIT.
Tuesday, March 6
Active Imaging and Display at MERL: A Selective Overview
Ramesh Raskar, Mitsubishi Electric Research Labs (MERL)
The processing for image analysis, display and user interaction can be significantly simplified by actively modifying and sensing scene illumination, camera parameters and display configuration. I will present an overview of several projects at MERL in Computational Photography, Active Illumination for scanning and Projector-based Displays.
Ramesh Raskar joined MERL as a Research Scientist in 2000 after his doctoral research at U. of North Carolina at Chapel Hill. His work spans a range of topics in computer vision and graphics including computational photography, projective geometry, non-photorealistic rendering and intelligent user interfaces. Current projects include flutter shutter camera, composite RFID (RFIG), multi-flash non-photorealistic camera for depth edge detection, locale-aware mobile projectors, high dynamic range video, image fusion for context enhancement and quadric transfer methods for multi-projector curved screen displays.
Dr. Raskar received the TR100 Award, Technology Review's 100 Top Young Innovators Under 35, 2004, Global Indus Technovator Award 2003, instituted at MIT to recognize the top 20 Indian technology innovators on the globe, Mitsubishi Electric Valuable Invention Award 2004 and Mitsubishi Electric Information Technology R&D Award 2003. He is a member of the ACM and IEEE.
Homepage: http://www.merl.com/people/raskar/raskar.html
Tuesday, March 13
High-Resolution, Real-Time Geometry Video Acquisition
Using a Phase-Shifting Method
Song Zhang, Harvard University
High-resolution, real-time 3D geometric shape measurement for dynamically deformable objects has a huge potential for applications in many areas, including entertainment, medical, design and manufacturing, etc. However, due to the challenging nature of the problem, no other system with such capability has ever been developed. In this talk I will discuss a recently developed system with such capabilities.
The system we develop is based on a digital fringe projection and phase-shifting technique. It utilizes a single-chip Digital-Light-Processing (DLP) projector to project computer generated fringe patterns onto the object and a high-speed Charge-Coupled-Device (CCD) camera synchronized with the projector to acquire the fringe images at a frame rate of 180 frames per second. Based on a three-step phase-shifting technique, each frame of the 3D shape is reconstructed using three consecutive fringe images. Therefore the 3D data acquisition speed of the system is 60 frames per second (faster than video speed 24fps). Together with fast 3D reconstruction algorithms and parallel processing software we developed, high-resolution, real-time 3D shape measurement is realized at a frame rate of up to 30 frames per second and a resolution of 300K points per frame. The motion of geometric shape changes, such as facial expressions can be accurately measurement with such a system.
Homepage: http://www.math.harvard.edu/~songzhang
Reading: Realtime Shape Measurement, Calibration of Structured Light Systems
Tuesday, March 20
Reflectance Modeling for Vision and Graphics
Todd Zickler, EECS, Harvard University
An image depends on a number of scene properties, including shape, reflectance, illumination and viewpoint. These properties interact to create specular highlights and other intricate visual effects. Understanding and modeling these effects is important for computer vision systems to succeed in real-world environments and for graphics systems to accurately synthesize visual appearance. In the past, the dominant paradigm in computer vision has been to assume that surfaces are purely diffuse (Lambertian) and either ignore complex reflectance effects or treat them as noise. This talk is meant to convey an alternative approach. Although our visual world contains a vast collection of different materials, there are common physical reflectance properties (reciprocity, isotropy, separability, spatial coherence, etc.) that are shared by broad classes of materials. By developing computational tools that exploit these properties, we can use image data more efficiently and improve many vision and graphics systems. To demonstrate this approach, I present tools that can be used for a variety of applications, including 3D reconstruction, modeling, recognition, tracking, and segmentation. Bio: Todd Zickler received a B. Eng. degree in honors electrical engineering from McGill University in 1996, and an MS degree in electrical engineering from Yale University in 2001. He received a PhD degree in electrical engineering from Yale in 2004, at which point he joined Harvard University as an assistant professor of electrical engineering in the Division of Engineering and Applied Sciences. His research interests span computer vision, image processing and computer graphics, and he is currently focused on image-based modeling and efficient representations for visual appearance. In 2006, he was the recipient of an NSF Career Award titled "Foundations for Ubiquitous Image-based Appearance Capture."
Homepage: http://www.eecs.harvard.edu/~zickler
Tuesday, April 3
3D Scanning at the Yale Graphics Group
Holly Rushmeier, CS, Yale University
Homepage: http://graphics.cs.yale.edu/holly/
Tuesday, April 17
Architecture - Archaeology - Virtual Reality - Virtual Heritage: Rescue the past. Save the world
Donald Sanders, Institute for the Visualization of History
Homepage: http://www.vizin.org/