This paper presents a new deformable modeling strategy aimed at integrating shape and appearance in a unified space. If we think traditional deformable models as "active cont...
We present a geometric model and a computational method for segmentation of images with missing boundaries. In many situations, the human visual system fills in missing gaps in ed...
—We present a fully automatic face recognition algorithm and demonstrate its performance on the FRGC v2.0 data. Our algorithm is multimodal (2D and 3D) and performs hybrid (featu...
A new adaptive multispectral image compression technique based on the regions identified is proposed. The algorithm is adaptive in the sense that according to the data type class ...