New Advances in Computer Vision and Graphics

Zhengyou Zhang


In this talk, I will briefly present four pieces of my recent work with my collaborators at Microsoft Research.

The first is a new 3D reconstruction technique called model-based bundle adjustment, and has been applied to face modeling. It was presented at ICCV2001.

The second is a novel technique based on ratio images to map one person's facial expression to others. It is a simple but very effective technique because it can transfer all the facial details due to facial expression. It is presented at Siggraph 2001.

The third is called Visual Screen. With an ordinary camera, we transform an ordinary monitor into a touch screen.

The fourth is called Visual Panel. With an ordinary camera, we transform a rectangluar panel (e.g., a piece of paper) into a virtual mouse, a virtual keyboard and a virtual joystick.

Maintained by Philippos Mordohai