Visual Working Memory

Abstract

Visual working memory (VWM) is an umbrella term for processes of short-term retention and manipulation of visual information which support performance in an ongoing task. In the last decade studies in the field resulted in numerous revolutionary, both behavioural and neuropsychological, findings that have changed our understanding of VWM. In this review, we discuss some of the most important theoretical ideas and findings that have defined research orientations in the field. We begin this paper with placing VWM in the context of similar cognitive processes. After describing the most commonly used tasks in the studies of VWM, in the body of this paper we are reviewing two dominant theories of VWM - slot models and resource models of VWM. After almost forty years of empirical support of slot models, recent studies conducted under resource models paradigm are reappraising our knowledge about VWM. In order to conclude the section on VWM architecture, we review studies that aimed to answer the question whether the VWM representations are comprised of specific visual features or already integrated objects. Since the two opposing approaches cannot completely explain existing findings, the most promising approach is the one which suggests a hierarchical model of encoding, both at the level of features and integrated objects. In the second part of the paper, we review three main tasks of VWM: encoding, retrieval, and forgetting of the information. Fields of studying these components are marked with verification of clearly opposing models, along with methodological innovations proposed in order to thoroughly examine those models. Here, we review them with an aim to elucidate the nature of core VWM components.