The face-space duality hypothesis: a computational model
This work addresses a gap in psychological models for dynamic facial features, offering a unified computational approach for researchers in face recognition and cognitive science.
The authors tackled the problem of unifying identity and expression representations in face recognition by proposing a computational model based on the face-space duality hypothesis, which was validated through mathematical presentation and experiments showing it supports both recognition tasks.
Valentine's face-space suggests that faces are represented in a psychological multidimensional space according to their perceived properties. However, the proposed framework was initially designed as an account of invariant facial features only, and explanations for dynamic features representation were neglected. In this paper we propose, develop and evaluate a computational model for a twofold structure of the face-space, able to unify both identity and expression representations in a single implemented model. To capture both invariant and dynamic facial features we introduce the face-space duality hypothesis and subsequently validate it through a mathematical presentation using a general approach to dimensionality reduction. Two experiments with real facial images show that the proposed face-space: (1) supports both identity and expression recognition, and (2) has a twofold structure anticipated by our formal argument.