Holistic face processing across neural and behavioral measures

Abstract

Holistic face processing is a fundamental aspect of human visual perception, distinguishing face processing from other visual stimuli processing. It involves the integration of various facial features into a whole face percept. This thesis investigates the cognitive and neural mechanisms underlying holistic face processing through a series of behavioral and electrophysiological experiments. Using event-related potentials (ERPs), this research explores the effects of stimulus orientation and fixation location on face and object processing. In addition, using accuracy and reaction time measures, the thesis explores the effects of face size on face recognition and gender discrimination. The first two ERP studies focus on the face inversion effect and its modulation by fixation location for faces and houses. A gaze-contingent paradigm ensures precise control over where participants fixate on upright and inverted faces and houses. Results reveal that fixation location significantly influences ERP responses, particularly the N170 component, a key marker of early face processing, but the strongest effects were seen at earlier and later time periods. Notably, holistic processing appears flexible, with fixation on specific facial features exerting differential effects on neural activity. Importantly, we found a strong house inversion effect and similar modulations of the neural activity to houses with fixation location. The findings challenge the traditional view of holistic processing as a rigid, fixation-independent mechanism only at play for faces and not objects. A third behavioural study examines the role of face size in recognition accuracy and reaction times across orientations and tasks. Behavioral results indicate that holistic face processing varies with size, orientation, and task with optimal processing occurring within a specific range between extreme face sizes. These results provide new insights into how face size modulates perception and recognition efficiency. Together, these studies demonstrate that holistic face processing is dynamic and influenced by multiple factors, including fixation location, orientation, and size. By employing advanced statistical techniques such as mass univariate analysis of ERP data, this research provides a more nuanced understanding of the spatiotemporal dynamics of face perception. The findings have implications for models of face recognition, cognitive neuroscience, and applications in artificial intelligence and biometric technologies.