Exploring the Animacy Dimension: Dissecting the Effects of Visuo-Semantic Clash in Object Recognition

Abstract

Object recognition is a multifaceted process that involves detecting and distinguishing visual features while processing their semantic associations. The aim of this thesis is to explore the animacy dimension with behavioral measures. By assessing the human ability to discriminate between animate and inanimate entities when their appearance stands in contrast with their identity, we hope to gain a more thorough understanding of the effects of semantic representations on high-level vision. Moreover, we investigate whether visual properties truly are the most important cue in object recognition or if semantic connotations might play an essential part in the recognition of objects. Our results suggest that visual features dominate perception when stimuli belong to their true object category (e.g., inanimate-looking objects). However, when an entity’s appearance stands in contrast with its object category (e.g., animate-looking objects), semantic representations seem to interfere. Additionally, animals and animate-looking objects seem to be processed differently than inanimate-looking objects, suggesting a specific visual animacy space.