Actas de congresos
The Best Of Both Worlds: Combining 3d Deformable Models With Active Shape Models
Registro en:
Proceedings Of The Ieee International Conference On Computer Vision. , v. , n. , p. - , 2007.
10.1109/ICCV.2007.4409015
2-s2.0-50649094502
Autor
Vogler C.
Li Z.
Kanaujia A.
Goldenstein S.
Metaxas D.
Institución
Resumen
Reliable 3D tracking is still a difficult task. Most parametrized 3D deformable models rely on the accurate extraction of image features for updating their parameters, and are prone to failures when the underlying feature distribution assumptions are invalid. Active Shape Models (ASMs), on the other hand, are based on learning, and thus require fewer reliable local image features than parametrized 3D models, but fail easily when they encounter a situation for which they were not trained. In this paper, we develop an integrated framework that combines the strengths of both 3D deformable models and ASMs. The 3D model governs the overall shape, orientation and location, and provides the basis for statistical inference on both the image features and the parameters. The ASMs, in contrast, provide the majority of reliable 2D image features over time, and aid in recovering from drift and total occlusions. The framework dynamically selects among different ASMs to compensate for large viewpoint changes due to head rotations. This integration allows the robust tracking of faces and the estimation of both their rigid and nonrigid motions. We demonstrate the strength of the framework in experiments that include automated 3D model fitting and facial expression tracking for a variety of applications, including sign language. ©2007 IEEE.
Beymer, D., Poggio, T., Image representations for visual learning (1996) Science, 272 (5270), pp. 1905-1909 Blake, A., Isard, M., (1998) Active Contours: The Application of Techniques from Graphics, Vision, Control Theory and Statistics to Visual Tracking of Shapes in Motion, , Springer-Verlag Blanz, V., Vetter, T., A morphable model for the synthesis of 3D faces (1999) SIGGRAPH, pp. 187-194 Brand, M., Morphable 3D models from video (2001) CVPR Cootes, T., Taylor, C., Active shape models - their training and application (1995) CVIU, 61 (1), pp. 38-59 de Carlo, D., Metaxas, D., Optical flow constraints on deformable models with applications to face tracking (2000) IJCV, 38 (2), pp. 99-127. , July Dimitrijevic, M., Ilic, S., Fua, P., Accurate face models from uncalibrated and ill-lit video sequences (2004) CVPR, pp. 1034-1041 Goldenstein, S., Vogler, C., When occlusions are outliers (2006) Workshop on the 25 Years of RANSAC (in conjunction with CVPR) Goldenstein, S., Vogler, C., Metaxas, D., Statistical Cue Integration in DAG Deformable Models (2003) PAMI, 25 (7), pp. 801-813 Goldenstein, S., Vogler, C., Metaxas, D., 3D facial tracking from corrupted movie sequences (2004) CVPR Goodall, C., Procrustes methods in the statistical analysis of shape (1991) Journal of the Royal Statistical Society B, 53, pp. 285-339 Isard, M., Blake, A., (1998) Condensation: Conditional density propagation for visual tracking, 29 (1), pp. 5-28 Jones, M., Poggio, T., Multidimensional morphable models (1998) ICCV, pp. 683-688 Lanitis, A., Taylor, C.J., Cootes, T.F., Automatic interpretation and coding of face images using flexible models (1997) PAMI, 19 (7), pp. 743-756 Pighin, F., Szeliski, R., Salesin, D., Modeling and animating realistic faces from images (2002) IJCV, 50 (2), pp. 143-169 Tao, H., Huang, T., Visual Estimation and Compression of Facial Motion Parameters: Elements of a 3D Model-Based Video Coding System (2002) IJCV, 50 (2), pp. 111-125 Cootes, T.F., Edwards, G., Taylor, C.J., Active appearance models (2001) PAMI, 23 (6), pp. 681-685 Viola, P., Jones, M., Robust real-time face detection (2004) IJCV, 57 (2), pp. 137-154 Vogler, C., Goldenstein, S., Stolfi, J., Pavlovic, V., Metaxas, D., Outlier rejection in high-dimensional deformable models (2007) IVC, 25 (3), pp. 274-284 Wang, L., Hu, W., Tan, T., Face tracking using motion-guided dynamic template matching (2002) ACCV Xiao, J., Baker, S., Matthews, I., Kanade, T., (2004) Real-time combined 2d+3d active appearance models, , CVPR