Scene perception: A failure to find a benefit from prior expectancy or familiarity.

Abstract

In our everyday world, we typically have an expectancy as to the kinds of scenes that we will see from one glance to the next. Also, many of the scenes that we do see are familiar in the sense that they have been experienced before. Do these factors influence the perception of a scene? In three experiments, priming subjects with a verbal descriptor of a scene was not found to improve reliably the perception of that scene as assessed by the speed and accuracy of detecting an incongruity between an object and its setting (Experiments 1 and 2) or a specified target object (Experiment 3). Also, in attempting to perceive these scenes, subjects could not capitalize on the residue from prior exposures of a scene's background, even though those backgrounds had been processed to the point where semantic information had been extracted from them. Although these results are inconsistent with recent speculations on the role of frames in scene perception, recent experiments on the perception of a scene from a single fixation, and film-editing practice with "flash cuts." The implications of these results are that the mechanisms for perceiving and interpreting nondegraded real-world scenes are so quick and efficient that conditions can readily be found in which priming and prior exposures of substantial portions of scenes are not helpful for perceiving and judging certain aspects of those scenes.

Effects of early common features on form perception

“…However, claims about top-down effects have often taken a strong form, in terms of mechanisms such as knowledgedriven hypothesis testing. For example, one hypothesis was that exposure to a scene's name (the semanticlevel) can expedite visual recognition of objects within the scene (see, e.g., Biederman, Teitelbaum, & Mezzanotte, 1983; see also, e.g., Henderson, 1977;Rumelhart, 1977). Most experiments have not been consistent with top-down hypotheses (see, e.g., Biederman et al, 1983;Johnston, 1978); consequently, few current models include topdown mechanisms.…”

Time course of object identification: Evidence for a global-to-local contingency.

“…However, the idea typically has been instantiated in terms of fairly strong mechanisms, such as top-down effects of prior knowledge and interactions between distinct processing levels (e.g., Henderson, 1977;Neisser, 1967;Rumelhart, 1977). For example, one hypothesis was that exposure to a scene's name, semantic-level identification, expedites visual-level identification of objects within the scene (e.g., Biederman, Teitelbaum, & Mezzanotte, 1983). Most of the data are inconsistent with strong top-down hypotheses (e.g., Biederman et al, 1983, Johnston, 1978, and few current models include strong topdown mechanisms.…”

“…For example, one hypothesis was that exposure to a scene's name, semantic-level identification, expedites visual-level identification of objects within the scene (e.g., Biederman, Teitelbaum, & Mezzanotte, 1983). Most of the data are inconsistent with strong top-down hypotheses (e.g., Biederman et al, 1983, Johnston, 1978, and few current models include strong topdown mechanisms. However, this leaves a largely unexplored middle ground between the strictly bottom-up and strong top-down approaches.…”

Quantitative modelling of perceptual salience at human eye position

“…Hence, we do not expect perfect agreement between model-predicted salience and human eye position. In particular, our bottom-up model as used here does not yet account, among others, for how the rapid identification of the gist (semantic category) of a scene may provide contextual priors to more efficiently guide attention towards target objects of interest (Biederman, Teitelbaum, & Mezzanotte, 1983;Friedman, 1979;Hollingworth & Henderson, 1998;Oliva & Schyns, 1997;Potter & Levy, 1969;Torralba, 2003); how search for a specific target might be guided top-down, for example by boosting visual neurons tuned to the attributes of the target (Ito & Gilbert, 1999;Moran & Desimone, 1985;Motter, 1994;Müller, Reimann, & Krummenacher, 2003;Reynolds, Pasternak, & Desimone, 2000;Treue & Maunsell, 1996;Treue & Trujillo, 1999;Wolfe, 1994Wolfe, , 1998Wolfe, Cave, & Franzel, 1989;Yeshurun & Carrasco, 1998); or how task, expertise, and internal scene models may influence eye movements (Henderson & Hollingworth, 2003;Moreno, Reina, Luis, & Sabido, 2002;Nodine & Krupinski, 1998;Noton & Stark, 1971;Peebles & Cheng, 2003;Savelsbergh, Williams, van der Kamp, & Ward, 2002;Tanenhaus, Spivey-Knowlton, Eberhard, & Sedivy, 1995;Yarbus, 1967). Nevertheless, our hypothesis for this study is that a more realistic simulation framework might yield better agreement between human and model than a less realistic one.…”