Improved Facial Mapping and Modeling

 

Executive Summary

 

Improving facial recognition and facial mapping is an area of research which has been rapidly advancing in the last decade in order to improve both security and entertainment products. This MSU developed technology has the ability to improve both high-quality 3D face fitting and facial recognition technologies allowing for the improvement of video identification and tracking, computer generated models, and face swapping software. In addition to increasing the resolution of mapping, this software has the ability to recognize expressions of a facial image.

 

 

Description of Technology

 

The improvements in quality are facilitated by the increase in the amount and selection of landmark points on a person’s face when generating a 3D image from a source image. This technology is able to better distinguish the face from the image background. The density of landmark points is increased in areas where fine details occur, such as wrinkles or scars. When these improvements are combined, the resulting generated image is far more detailed than that produced using current technologies.

 

Key Benefits

  • Improved modeling of facial features
  • Trained with multiple databases, removes database bias

 

Applications

  • Augmented Reality/Virtual Reality
  • Facial Recognition
  • Computer Generated Images
  • Video Games

 

IP Status: Copyright

 

Licensing Rights Available

 

Non-exclusive rights

 

Inventors: Xiaoming Liu, Yaojie Liu, Amin Jourabloo

 

Tech ID: TEC2017-0086

 

 

 

Patent Information:

For Information, Contact:

Raymond Devito
Technology Manager
Michigan State University
devitora@msu.edu
Keywords: