MotionVis: Analyzing the Structure of a Motion Capture Database

The film and video game industries make use of large motion capture databases for creating realistic animations of human motion. Recently, significant research efforts have been devoted to data-driven animation techniques. These techniques allow new, realistic motion sequences to be automatically synthesized from existing motion capture sequences that are similar to the desired motion sequence. Unfortunately, identifying similar motion sequences is a challenging problem since logically similar motion sequences are often not numerically similar.

In this project, I developed an interactive visualization environment for analyzing the structure that is imposed on a human motion database by a given similarity metric. This environment provides insights into similarity metrics that are difficult to obtain by analyzing numerical results or by using existing visualization environments.


Information related to this project:

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License