Information maximization in face processing

Abstract

This perspective paper explores principles of unsupervised learning and how they relate to face recognition. Dependency coding and information maximization appear to be central principles in neural coding early in the visual system. These principles may be relevant to how we think about higher visual processes such as face recognition as well. The paper first reviews examples of dependency learning in biological vision, along with principles of optimal information transfer and information maximization. Next, we examine algorithms for face recognition by computer from a perspective of information maximization. The eigenface approach can be considered as an unsupervised system that learns the first-and second-order dependencies among face image pixels. Eigenfaces maximize information transfer only in the case where the input distributions are Gaussian. Independent component analysis (ICA) learns high-order dependencies in addition to first-and second-order relations, and maximizes information transfer for a more general set of input distributions. Face representations based on ICA gave better recognition performance than eigenfaces, supporting the theory that information maximization is a good strategy for high level visual functions such as face recognition. Finally, we review perceptual studies suggesting that dependency learning is relevant to human face perception as well, and present an information maximization account of perceptual effects such as the atypicality bias, and face adaptation aftereffects. r

Internal representations for face detection: An application of noise‐based image classification to BOLD responses

“…Conversely, these previous findings support the idea that the most robustly encoded features are those most diagnostic about faces as a class. In this sense, face encoding appears to reflect the objective structure and statistics of face images [Bartlett, ] in a manner that is similar to the way early visual representations reflect the low‐level statistics of natural images [Barlow, 1961; Olshausen and Field, ].…”

Visual adaptation and face perception

“…This theoretical approach has led to important insights into early visual coding and may similarly reveal the principles involved in high-level representations. An example of how these analyses have been applied to the specific case of face processing can be found in Bartlett [ 158 ]. Here, we consider how optimal coding schemes derived from the analyses of colour vision can be used to predict how faces might be processed.…”

Adaptation and visual coding

“…An interesting example is in face perception, which like many other patterns shows only mixed evidence for improved discrimination with adaptation (Ng, Boynton, & Fine, 2008; Rhodes, Maloney, Turner, & Ewing, 2007; Rhodes, Watson, Jeffery, & Clifford, 2010). The distribution of faces is again peaked around the average, and like luminance or chromatic variations, this might predict a sigmoidal response to optimally encode the stimulus differences (Bartlett, 2007). Yet, discrimination and adaptation have instead been found to be consistent with a largely linear response function for facial configurations, even over ranges larger than faces naturally vary (Susilo, McKone, & Edwards, 2010).…”