Title :
Facial expression synthesis
Author_Institution :
Sch. of Eng., Flinders Univ. of South Australia, Bedford Park, SA, Australia
fDate :
6/21/1905 12:00:00 AM
Abstract :
A new method is presented in this paper for synthesis of an expressionless face image from a face image containing an arbitrary known expression. The proposed method enhances the performance of the existing principal components analysis which can be trained using a limited set of expressionless face images. This performance enhancement is achieved through the extraction of the basis set of the example face images in a global-local-based fashion
Keywords :
face recognition; gesture recognition; principal component analysis; visual databases; arbitrary known expression; expressionless face image synthesis; global-local-based method; principal components analysis; Australia; Face recognition; Image analysis; Image databases; Image generation; Image recognition; Information analysis; Performance analysis; Principal component analysis;
Conference_Titel :
Image Processing, 1999. ICIP 99. Proceedings. 1999 International Conference on
Conference_Location :
Kobe
Print_ISBN :
0-7803-5467-2
DOI :
10.1109/ICIP.1999.821713