84
Voted
CVPR
2004
IEEE
16 years 24 days ago
2004
IEEE
In this paper we perform 3D face tracking on corrupted video sequences. We use a deformable model, combined with a predictive filter, to recover both the rigid transformations and...