A feature-based approach for individualized human head modeling

Published in The Visual Computer, 2001

Abstract: We present a new feature-based approach to efficiently model individualized human heads. First, we generate a generic head model from a discrete data set using a displaced butterfly subdivision scheme. Our generic model describing fine details on the human head is feature-based and semi-regular. We represent our generic model using a feature mesh together with a hierarchical detail set. To individualize the head model, we deform the feature mesh by adjusting a set of prescribed feature points; we then add the detail set back to synthesize a smooth head model for individuals. We show that using our technique we can achieve great efficiency both in highly real- istic head modeling and in a wide range of downstream applications.

Download paper here

More information

Recommended citation: Yong-Jin Liu, Ming-Fai Yuen, Shan Xiong (2002) A feature-based approach for individualized human head modeling. The Visual Computer, Vol. 18, No. 5-6, pp. 368-381, 2002.