Tutorial Description
A computer vision based morph between 6 hot celebrities, and 6 random women, (perhaps on their way to becoming celebrities themselves). The vision system we implemented in Matlab optimized the sequence of faces in order to generate the smoothest morphs face to face without manually entering control point pairs. The sequence optimizations were implemented using Simulated Annealing to improve a Greedy Tour as an approximation to the NP Complete Hamiltonian Path Problem. You can tell that the system is wise, because Lindsay Lohan is morphed into Dar Williams, producing a few frames of what could only be achieved through decades of expensive cloning procedures. This is a Final Exam Team Project for the University of Pennsylvanias GRASP Robotics Lab course: CIS-581, Computer Vision. Taught by Professor Jianbo Shi, Fall 2008. (highly recommended) Authors: Srinivasan Dwarakanathan, Hua-Min Shen, Jonathan Schmid.
Automatically Generated Morph Sequence - Video
No comments:
Post a Comment