faceshift 1.2 is packed with new features that improve data quality, accelerate the shooting workflow, and streamline post-processing. Our team has a strong academic background. We graduated from the three most renowned Swiss universities, ETH, EPFL, and University of Basel, and all our members have spent time in research labs around the world. We have published in the top conferences and journals, and are continuing to contribute to the academic community.

- Setup
For PrimeSense sensors … We now support Hi-Res Mode. This creates a smaller capture area but much better image quality. We’ve added Manual Texture Calibration, allowing pixel-perfect alignment between the image and depth sensors. This can dramatically increase quality on some sensors. Auto Exposure and White Balance can improve the video image considerably – making tracking even better.
- Training
We’ve made some under-the-hood improvements in Training. The new Training Model provides more natural expressions and better tracking. And no more searching for the appropriate profile to go with that performance you captured some time ago; faceshift 1.2 now saves the profile with the performances.
- Tracking
Tracking now has some features to make shooting and managing larger data sets easier. The Takes feature allows users to define a filename prefix for each performance. Very handy if you’re working on multiple projects, sequences, or scenes and want to organize the naming automatically. The Trim feature is a quick visual method to crop frame ranges. Then play these back at specific Playback Rates. Annotating is much improved; drag a feature point on a frame you actually teach the computer vision algorithm what to look for across the whole take! You’ll see the progress bar working as the features are quickly re-tracked across all frames.

Home:

www.faceshift.com

Huge Mirror for V.I.P Members