Some updates on my research.

As the video shows, on the left hand side, I was holding my mobile Samsung GS2 both hands with my elbow on the table while capturing the facial expressions. The jittering was quite visible and very shaky.

However, on the right hand side, I made my own facial mocap "Fork" as you can see in the photo. It's basically 3 pieces of wood, attached together with some metal plate and that's it. The camera goes between as well as my head. Pretty simple and the results were stunning in comparison without using the DIY fork.

I hope this helps and explains how important Stability is needed while doing facial motion capture.

Next? Building my real helmet. The metal guy in the shop can build me the gig for $40 bucks!
I just have to come up with the helmet.

Facial Motion Capture - Jittering Comparison - YouTube