Viewpoint Invariant Face Recognition using Independent Component Analysis and Attractor Networks

By M. Bartlett, Terrence J. Sejnowski. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Conditional variance penalties and domain shift robustness (2017, background): “In the computer vision literature, various works have used identity information to achieve pose invariance in the context of face recognition (Bartlett and Sejnowski, 1996; Tran et al., 2017).”
Learning Optimized Features for Hierarchical Models of Invariant Object Recognition (2003, background): “Bartlett and Sejnowski (1997) showed that for face recognition, ICA representations have advantages over representations based on principal component analysis (PCA) with regard to pose invariance and classification performance.”
Recognizing faces with PCA and ICA (2003, methodology): “Bartlett, et al. [4] [6], Liu and Wechsler [30], and Yuen and Lai [41] claim that ICA outperforms PCA for face recognition, while Baek et al. [1] claim that PCA outperforms ICA and Moghaddam [32] claims that there is no statistical difference in performance between the two.”
A Unifying Information-Theoretic Framework for Independent Component Analysis (2000, background): “Assumption 2 is the basis of ICA and can be expressed as follows: p(s(t)) = M Y i=1 p(si(t)) (6) For assumption 3 the unmixing of two Gaussian sources is ill posed when the sources are white random processes.”
Guarding Against Adversarial Domain Shifts with Counterfactual Regularization (2017, background): “ latent factors explicitly as in a generative framework. In the computer vision literature, various works have used identity information to achieve pose invariance in the context of face recognition (Bartlett and Sejnowski, 1996; Tran et al., 2017). More generally, the idea of exploiting various obs”
Kernel Eigenfaces vs. Kernel Fisherfaces: Face recognition using kernel methods (2002, background): “Though our experiments using ICA seem to contradict to the good empirical results reported in [4] [3] [2], a close look at the data sets reveals a signi cant di erence in pose and scale variations of the face images in the AT&T database, whereas a subset of frontal view FERET face images with change”