Proceedings of the 4th international conference on Computer graphics and interactive techniques in Australasia and Southeast Asia
A template-based approach to automatic face enhancement
Pattern Analysis & Applications
Face alignment by Explicit Shape Regression
CVPR '12 Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
Hi-index | 0.00 |
In this paper, we focus on the modification of facial shapes for the purpose of efficient face beautification. The pursuit of beautiful appearance for human greatly inspires related studies. We utilize Laplacian to characterize facial geometry and adjust triangular meshes neighboring facial contour in order to beautify faces while preserving major structures. The proposed method requires a small amount of local computations and no additional storage for training, which renders efficient implementation applicable for mobile devices. We develop the system on an intelligent mobile platform and the performance is validated by a survey on 70 facial images since no objective metric for beauty exists. Twenty volunteers are invited to grade the results labeled as level 1, 2 and 3 that stand for worse, unchanged and better, respectively. The results show that 81.50% of inputs are believed becoming more beautiful, and another 10.07% of them are considered unchanged.