Face recognition robust to head pose from one sample image

Most face recognition systems only work well under quite constrained environments. In particular, the illumination conditions, facial expressions and head pose must be tightly controlled for good recognition performance. In 2004, we proposed a new face recognition algorithm, Adaptive Principal Component Analysis (APCA) [4], which performs well against both lighting variation and expression change. But like other eigenface-derived face recognition algorithms, APCA only performs well with frontal face images. The work presented in this paper is an extension of our previous work to also accommodate variations in head pose. Following the approach of Cootes et al, we develop a face model and a rotation model which can be used to interpret facial features and synthesize realistic frontal face images when given a single novel face image. We use a Viola-Jones based face detector to detect the face in real-time and thus solve the initialization problem for our Active Appearance Model search. Experiments show that our approach can achieve good recognition rates on face images across a wide range of head poses. Indeed recognition rates are improved by up to a factor of 5 compared to standard PCA.

Reflexive optokinetic nystagmus in younger and older observers under photopic and mesopic viewing conditions

PURPOSE. To investigate the effect of age on optokinetic nystagmus (OKN) in response to stimuli designed to preferentially stimulate the M-pathway. METHOD. OKN was recorded in 10 younger (32.3 +/- 5.98 years) and 10 older (65.6 +/- 6.53) subjects with normal vision. Vertical gratings of 0.43 or 1.08 cpd drifting at 5 degrees/s or 20 degrees/s and presented at either 8% or 80% contrast were displayed on a large screen as full-field stimulation, central stimulation within a central Gaussian-blurred window of 15 diameter, or peripheral stimulation outside this window. All conditions apart from the high-contrast condition were presented in a random order at two light levels, mesopic (1.8 cdm(-2)) and photopic (71.5 cdm(-2)). RESULTS. Partial-field data indicated that central stimulation, mesopic light levels, and lower temporal frequency each significantly increased slow-phase velocity (SPV). Although there was no overall difference between groups for partial-field stimulation, full-field stimulation, or low-contrast stimulation, a change in illumination revealed a significant interaction with age: there was a larger decrease in SPV going from photopic to mesopic conditions for the older group than the younger group, especially for higher temporal frequency stimulation. CONCLUSIONS. OKN becomes reflexive in conditions conducive to M-pathway stimulation, and this rOKN response is significantly diminished in older healthy adults than in younger healthy adults, indicative of decreased M-pathway sensitivity.