Video

A demonstration of Serendip’s RankViewer. Serendip is a corpus exploration tool that uses probabilistic topic modeling as a lens through which to view documents. The demonstration uses a topic model generated on a corpus of 1,080 digitized English texts published between 1530 and 1799.

Video from our 2011 Transactions on Graphics Paper.

This system provided a practical approach to warp stabilization. It is the key technical piece that lead to Adobe’s Warp Stabilizer.

This work was done with Feng Liu, Michael Gleicher, Jue Wang, Hailin Jin, and Aseem Agarwala.

This is a video demonstration from our SIGGRAPH 2009 paper “Content-Preserving Warps for 3D Video Stabilization”. This is an early predecessor to the technology that became Adobe’s Warp Stabilizer (although, that is really based on a later paper).

This video demonstrates Re-Cinematography - our approach for post-process video stabilization that tries to produce high-quality camera motions. The original paper (at MultiMedia 2007) won the “best in track award,” and was invited to be extended as a journal paper in TOMCCAP.

Note: this is taken from a “FLV” file so, at one point it was highly compressed using 2007 technology. It was before we had reliable H.264 (or better).

We present an example-based motion synthesis technique that generates continuous streams of high-fidelity, controllable motion for interactive applications, such as video games.

We present automated system called Virtual Videography that employs the art of videography to mimic videographer-produced lecture videos, while being unobtrusive when recording.

The system is described in a 2007 Paper.

We present a simple and efficient technique for synthesizing high-fidelity motions by attaching, or splicing,the upper-body action of one motion example to the lower-body locomotion of another.