On Transforming Statistical Models for Non-Frontal Face Verification. Sanderson, C., Bengio, S., & Gao, Y. Pattern Recognition, 39(2):288–302, 2006.
On Transforming Statistical Models for Non-Frontal Face Verification [link]Paper  abstract   bibtex   
We address the pose mismatch problem which can occur in face verification systems that have only a single (frontal) face image available for training. In the framework of a Bayesian classifier based on mixtures of gaussians, the problem is tackled through extending each frontal face model with artificially synthesized models for non-frontal views. The synthesis methods are based on several implementations of Maximum Likelihood Linear Regression (MLLR), as well as standard multi-variate linear regression (LinReg). All synthesis techniques rely on prior information and learn how face models for the frontal view are related to face models for non-frontal views. The synthesis and extension approach is evaluated by applying it to two face verification systems: a holistic system (based on PCA-derived features) and a local feature system (based on DCT-derived features). Experiments on the FERET database suggest that for the holistic system, the LinReg based technique is more suited than the MLLR based techniques; for the local feature system, the results show that synthesis via a new MLLR implementation obtains better performance than synthesis based on traditional MLLR. The results further suggest that extending frontal models considerably reduces errors. It is also shown that the local feature system is less affected by view changes than the holistic system; this can be attributed to the parts based representation of the face, and, due to the classifier based on mixtures of gaussians, the lack of constraints on spatial relations between the face parts, allowing for deformations and movements of face areas.
@article{sanderson:2006:pr,
  author = {C. Sanderson and S. Bengio and Y. Gao},
  title = {On Transforming Statistical Models for Non-Frontal Face Verification},
  journal = {Pattern Recognition},
  volume = 39,
  number = 2,
  pages = {288--302},
  year = 2006,
  url = {publications/ps/sanderson_2006_pr.ps.gz},
  pdf = {publications/pdf/sanderson_2006_pr.pdf},
  djvu = {publications/djvu/sanderson_2006_pr.djvu},
  original = {2006/non_frontal_pr},
  topics = {biometric_authentication},
  web = {http://dx.doi.org/10.1016/j.patcog.2005.07.001},
  abstract = {We address the pose mismatch problem which can occur in face verification systems that have only a single (frontal) face image available for training. In the framework of a Bayesian classifier based on mixtures of gaussians, the problem is tackled through extending each frontal face model with artificially synthesized models for non-frontal views. The synthesis methods are based on several implementations of Maximum Likelihood Linear Regression (MLLR), as well as standard multi-variate linear regression (LinReg). All synthesis techniques rely on prior information and learn how face models for the frontal view are related to face models for non-frontal views. The synthesis and extension approach is evaluated by applying it to two face verification systems: a holistic system (based on PCA-derived features) and a local feature system (based on DCT-derived features). Experiments on the FERET database suggest that for the holistic system, the LinReg based technique is more suited than the MLLR based techniques; for the local feature system, the results show that synthesis via a new MLLR implementation obtains better performance than synthesis based on traditional MLLR. The results further suggest that extending frontal models considerably reduces errors. It is also shown that the local feature system is less affected by view changes than the holistic system; this can be attributed to the parts based representation of the face, and, due to the classifier based on mixtures of gaussians, the lack of constraints on spatial relations between the face parts, allowing for deformations and movements of face areas.},
  categorie = {A},
}
Downloads: 0