My thesis work is offset in an existing technique which adapts motion capture data to conform with new constraints while maintaining a style of motion likely in the physical realm. New constraints might specify a different landing position in a jumping motion -or all the in-betweens of two extreme landing positions. This is advantageous when simple interpolation fails to produce believable motion. 

My contribution in this relation is a balance objective that is more grounded in dynamics and enables believable motion transformations for a wider range of motion editing problems. This dynamic balance objective is offset in the 3D spring-loaded inverted pendulum model but also incorporates rotational dynamics.

In this project I dealt with subjects such as;
Rotational dynamics
Inverse kinematics (Cyclic Coordinate Descend)
Nurbs
Evolutionary Algorithms
Non-linear Constrained Optimization

more