[Japanese | Thesis | Researches in Minoh Lab | Minoh Lab]
This article discusses how to realize the process of converting the user's facial image into an avatar's facial image with preserving the expression. By realizing it, the user can control avatar's expression by changing his/her own expression. In this process, the avatar's facial image should be synthesized just as the user intented.
In this article, it is proposed to adapt the conversion from a real facial image into an avatar's, to the user's intention. This is realized by an example-based approach, in which the correspondence between the facial images of the user and the avatar is estimated from its examples obtained through the interactions with the user. With the increase of the examples, the conversion gradually becomes to reflect the user's intention.