An efficient facial image morphing algorithm based on the self-organizing feature map algorithm is developed. A simplified facial muscle model is also proposed to provide us the information of how to determine the neighborhood functions so as to achieve a more natural look of the warped images. The value of the parameter σ in the neighborhood function plays an important role of local influence zone of reference points. The selection of σ should depend on the size of facial image and the location of the anchor point.
|Number of pages||4|
|State||Published - 1999|
|Event||International Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, USA|
Duration: 10 Jul 1999 → 16 Jul 1999
|Conference||International Joint Conference on Neural Networks (IJCNN'99)|
|City||Washington, DC, USA|
|Period||10/07/99 → 16/07/99|