A Connectionist Model of MT+/Mstd Explains Human Heading Perception in the Presence of Moving Objects

摘要: Human heading perception is robust in the presence of moving objects, except when object crosses focus expansion (FoE). If approaches observer, shifted towards FoE [1]. maintains a fixed distance, away from [2]. This data has resulted theory that differential motion operators are used to determine heading. Primate neurophysiology, however, indicates additive MT cells which combine across pools sensitive V1, and not cells, assess dorsal MST [3]. The present work utilizes components ViSTARS (Visually-guided Steering, Tracking, Route Selection) Biologically Inspired Cognitive Architecture (BICA) analyze reconcile data. ViSTARS BICA relating [4] were analyzed minimal modifications made provide distance dependent weighting template match. results indicate required explain human bias, provided detailed model V1-MT-MSTd used. bias objects representation as use population codes represent spatial properties within system. Work ongoing fully characterize causes nature this bias. demonstrates utility only extending our computational capabilities but also providing an overall framework with understand on it built.