A Holistic Approach to Robust and Scalable Object Recognition

This project is conducted at Advanced Digital Sciences Center (ADSC), as part of the Interactive Digital Media (IDM) sub-program. The PI leading our research efforts is Dr. Yi Ma from UIUC and MSRA.


[Left plot] Training acquisition system: Four projectors and two cameras controlled by one computer. [Top-right plot] 38 traning images of one subject captured by the training acquisition system. [Bottom-right plot] Representative testing examples, including indoor/outdoor images (1st row), images of subjects with eyeglasses (2nd row), and image of subjects with sunglasses (3rd row).


The goal of this project is to investigate and develop principled computational methods for highly effective, scalable, and robust object recognition systems. The objects of interest to this project range from objects with specific identities such as human faces, (Chinese) characters, and city landmarks to more loosely defined object classes such as automobiles, pedestrians, or trees. Such object recognition systems are urgently needed by a broad range of modern applications: augmented reality; indexing, ranking, and searching of massive images and videos over the Internet; and security surveillance.