The Past Shapes the Present: How Visual Learning Guides Visual Perception and Memory

The visual world around us contains an overwhelming amount of information, yet our perceptual system can only process a fraction of it at any given moment, and even less is retained in working memory. To efficiently achieve our behavioral goals, we must optimize the effectiveness of visual perception and memory. Luckily, the visual environment around us is often stable and highly structured, characterized by recurring regularities (e.g., books typically appear on the desks rather than on the floor) and repeated stimuli (e.g., our own faces). This dissertation investigates whether and how visual learning affects visual perception and memory. Specifically, we looked at the effects of the learning of regularities (i.e., the extraction of distributional probabilities of stimuli in space or time, termed as ‘statistical learning’) and repeated stimuli (repetition of certain stimuli across time, termed as ‘visual repetition’) respectively. Chapter 1 demonstrated that statistical learning influences awareness, enabling faster conscious detection of frequently occurring visual objects compared to infrequent ones. Chapter 2 revealed that statistical learning enhances visual working memory performance, but only when the learned regularity is relevant to the task. Specifically, learning that an object frequently appears at a location does not improve memory recall, whereas learning that an object is frequently probed at a location does. Chapter 3 showed that passively viewed objects are encoded into working memory more efficiently when they later become task relevant. Chapter 4 further demonstrated that repeated exposure to target stimuli shifts observers' reliance from effortful short-term memory to long-term memory, thereby reducing the need for frequent (re)checks of the target stimuli. Overall, this dissertation showed that statistical learning optimizes the efficiency of perception and memory by prioritizing probable targets over improbable ones; in addition, visual repetition leads to the buildup of memory traces for future targets and an increased reliance on long-term memory. These findings improve our understanding of how past visual experiences shape visual perception and memory of the present.