Personal video sharing services such as YouTube have become popular because videos can easily be recorded in high-definition (HD) using a personal camcorder. However, it is difficult to broadcast an HD video via the Internet due to the large amount of data involved. We describe a method for generating videos with virtual camerawork based on object tracking technology. Once the user specifies the positions of the region of interest (ROI) on keyframes, our method can be used to generate virtual camerawork between two keyframes in a row based on the results of bi-directional tracking. We evaluated our method with subjective experiments that demonstrate its effectiveness.
