Week One: I spent the first week investigating previous papers that were published on similar research topics. This includes papers on free form mesh deformations, advanced rigging of non-human characters, and inverse kinematics. This research gave a good background of what problems have already been solved in computer graphics and what issues still need more research.

Week Two: I spent the week investigating Maya's retargeting capabilities. I used retargeting to animate oddly proportioned and non-human characters using Motion Capture Data. I learned a lot about the capabilities of Maya, but also its limitations. We also realized that our characters might not be driven only by skeletons, but also by splines and free form deformations.

Week Three: I researched advanced rigging on non human characters and providing these characters with their unique degrees of freedom that aren't driven by the human skeleton joints and the human's degrees of freedom. The goal of this specific rigging research is to have an animator rig a character as they would imagine it moving with the right degrees of freedom, then, take this skeleton and animate with a permutation of motion capture data.

Week Four: I am still working on advanced rigging of non-humanoid characters, but specifically focusing on IK (Inverse Kinematics) splines. Once applied to the character correctly, the IK Spline will interpolate, creating a smooth and sinuous motion of the character.

Week Five: I am working on integrating the functionality of IK Splines into the pre-existing motion capture data. I realize that the best way to accomplish this is to retarget the motion to a scaled down puppet skeleton, then use the puppet skeleton to run the non-humanoid skeleton.

Week Six: The problems I am still having with this approach mentioned above, is that as a user, in Maya, I can not choose whether I want one specific joint to influence the Control Vertices of the spline a certain percentage. Also, the spline can only be translated and constrained using Cluster Deformers, thus the spline and its control vertices will never lay exactly on the joint chain.

Week Seven: The method of integrating IK splines into Motion Capture data works well when a non-humanoid character is standing still and only moving its upper body. But, when the character locomotes, the IK splines portray some undefined behavior, which I haven't been able to fix or understand.

Week Eight: During our weekly meeting, we had the idea to make a scene come alive. Specifically, an outdoor scene would be more challenging, because the movement of many geological and biological objects are not previously defined. Objects that are man-made usually have specific joints and degrees of freedom. Some objects that could be interesting to animate are a seashell, clouds, different species of trees, shrubs, and flowers, mountain range, lava, streams, rocks, etc.

Week Nine: We had actors and actresses come in to the motion capture lab. We captured them performing as non-humanoid characters, including a chair, mountain, lava, clouds, and a tree. It was very interesting to watch the actors all conjure up a way to use their bodies to represent each object in their own unique way.

Week Ten: We processed the motion capture data for two different actor's performing as lava and started experimenting with how to animate it. My first approach was to skin a motion capture skeleton into a polygonal sphere and then cloak it in nCloth, so that it has a very smooth sinuous movement. The problem with this approach was that it dampened the actor's performance too much.

Week Eleven: Kate, the other DMP student, was working on understanding a different kind of deformation. She became an expert on lattices, and integrating lattices into motion capture data. We compared and contrasted our results for the two different ways of deforming and animating non human characters. We decided that lattices are better to use if the object is more rigid and mass heavy, while splines create more fluid and sinuous movement.

Week Twelve: We spent that last week identifying problems still existing that could be researched further; 1.) Collisions are not dealt with or even addressed. 2.)There is no way to preserve the motion and fix the collisions. 3.) How to decide what Degrees of Freedom to map from Human to Non-Humanoid character. 4.) How to deal with different actors using unique degrees of freedom while portraying the same character.