Overview of research
Some of the general themes of research in the group are listed below. For more details, please click on the projects link above.
Interactive geometric modeling
3D content creation is in high demand in the film and gaming industry, design and manufacturing and many more areas, and it is backed up by affordable 3D acquisition technologies. Yet, shape modeling tasks themselves, such as editing, deformation and animation, remain extremely laborious, requiring artistic skills and high technical expertise. Our research goal is to broaden the knowledge and understanding of shapes so as to empower the set of computerized modeling tools, and explore new ways to communicate the human intentions of shape manipulation to the computer in a natural and effective way.
Faculty members: Olga Sorkine, Denis Zorin
The task of recognizing objects in images or video is one that we do effortlessly, yet a consdierable faction of our brain is devoted to the task. Trying to give a computer the same ability has proven to be a challenging problem, currently the focus of much investigation. Our research focuses on applying a variety of methods from machine learning to the problem. These include: generative image representations; convolutional neural networks and other types of deep belief network.
Faculty members: Rob Fergus, Yann LeCun
Computational Photography aims to extend the capabilities of conventional photography through the use of computation. This permits the capture of enhanced or entirely novel representations of a scene. Our research in this area includes the use of low-level image statistics on blur removal problems; novel designs of camera hardware; and the acquisition and use of lightfields.