Face Alignment Demo

Leon Gu and Takeo Kanade

Upload your image here:

      - Faces must be frontal.
      - Images must be jpg format.
      - No unicode characters in the file name.
      - Minimal face size = 20x20 pixels.
This demo system detects faces and identifies detailed facial structures in images. The system is fully automatic; it is applicable to generic frontal faces with large appearance variations and 2D pose changes. This technique has a wide range of applications in face recognition, tracking, 3D reconstruction and animation. Check out the results on CMU/MIT face database, and the results on various images submitted by Internet users.
 

Related Publications:

Gu, L. and Kanade, T.
3D Alignment of Face in a Single Image
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, New York, 2006.
(pdf) (oral presentation)

Zhou, Y., Gu, L. and Zhang, H.
Bayesian Tangent Shape Model: Estimating Shape and Pose Parameters via Bayesian Inference
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Wisconsin, 2003.
(pdf) (oral presentation)




 
http://www.cs.cmu.edu/~gu
　






This demo for face detection and alignment is being made available for research purposes only. Before uploading your images, please make sure that the images are free for research or non-commercial use. Please note that most of images you might find on the Internet are copyrighted. If you are the owner or the creator of the images, by uploading them to our website you agree to license the images to CMU for research reuse under BSD License..