New version page

Psychological Science

Upgrade to remove ads

This preview shows page 1-2-3 out of 9 pages.

Save
View Full Document
Premium Document
Do you want full access? Go Premium and unlock all 9 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 9 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 9 pages.
Access to all documents
Download any document
Ad free experience

Upgrade to remove ads
Unformatted text preview:

http://pss.sagepub.com/Psychological Science http://pss.sagepub.com/content/early/2011/05/09/0956797611407930The online version of this article can be found at: DOI: 10.1177/0956797611407930 published online 9 May 2011Psychological ScienceKarla K. Evans, Todd S. Horowitz and Jeremy M. WolfePerceptionWhen Categories Collide : Accumulation of Information About Multiple Categories in Rapid Scene Published by: http://www.sagepublications.comOn behalf of: Association for Psychological Science can be found at:Psychological ScienceAdditional services and information for http://pss.sagepub.com/cgi/alertsEmail Alerts: http://pss.sagepub.com/subscriptionsSubscriptions: http://www.sagepub.com/journalsReprints.navReprints: http://www.sagepub.com/journalsPermissions.navPermissions: at Harvard Libraries on May 22, 2011pss.sagepub.comDownloaded fromPsychological ScienceXX(X) 1 –8© The Author(s) 2011Reprints and permission: sagepub.com/journalsPermissions.navDOI: 10.1177/0956797611407930http://pss.sagepub.comHumans can extract surprisingly complex semantic and statis-tical information from complex scenes that are presented very briefly or with limited focused attention. Within 200 ms, observers can assess the mean and distribution of size (Chong & Treisman, 2003), orientation (Parkes, Lund, Angelucci, Solomon, & Morgan, 2001), and other basic visual attributes (Chubb, Nam, Bindman, & Sperling, 2007; Melcher & Kowler, 1999) for arrays of many objects without needing to attend to each object. Within 120 ms, observers are still able to identify aspects of the meaning, or “gist,” of a novel scene (e.g., picnic or birthday party; Potter & Faulconer, 1975), as well as to rec-ognize small objects (Fei-Fei, Iyer, Koch, & Perona, 2007) or report their locations and spatial relations (Evans & Treisman, 2005; Tatler, Gilchrist, & Rusted, 2003). Even with shorter, masked viewing durations (19 to 50 ms), observers are able to classify a scene at basic (e.g., lake vs. forest) and superordi-nate (e.g., natural vs. urban) levels (Greene & Oliva, 2009; Joubert, Rousselet, Fize, & Fabre-Thorpe, 2007), and they can determine how pleasant it is (Kaplan, 1992). Furthermore, large objects can be identified (Thorpe, Fize, & Marlot, 1996; VanRullen & Thorpe, 2001), even when focused attention is engaged with another foveal task (Li, VanRullen, Koch, & Perona, 2002). We refer to properties that can be reported accurately under such circumstances as nonselective, because they can apparently be perceived without directing selective attention to individual objects (even though selective attention to individual objects is required by similar, seemingly simpler tasks; Vickery, King, & Jiang, 2005).The existence of such nonselective processing has inspired a set of feed-forward models of visual processing in which quite complex properties can be extracted from the initial sweep of neural activity from retina to cortex, without feed-back or other top-down influences (Fukushima & Miyake, 1982; Itti & Koch, 2001; Riesenhuber & Poggio, 2000; Rousselet, Thorpe, & Fabre-Thorpe, 2004; Serre, Oliva, & Poggio, 2007; Thorpe, Delorme, & Van Rullen, 2001). Some researchers have extrapolated from such models and results to the conclusion that the visual system automatically computes multiple nonselective properties at the same time (Rousselet et al., 2004; Serre et al., 2007). This conclusion comports well with the introspective impression that the visual world is rich and detailed, even in a single glance.However, this conclusion is not necessarily warranted by the data, and introspection is not definitive evidence. It is important to keep in mind that in most of the experiments just cited, observers were asked to report repeatedly on only one or two properties over many trials (e.g., Is there an animal in the scene?), and these properties were specified well in advance. One could explain the existing data equally well by assuming that only a small number of nonselective filters can be applied to the feed-forward visual data stream; although the number of Corresponding Author:Karla K. Evans, Harvard Medical School, Visual Attention Lab, 64 Sidney St., Suite 170, Cambridge, MA 02139 E-mail: [email protected] Categories Collide: Accumulation of Information About Multiple Categories in Rapid Scene PerceptionKarla K. Evans1,2, Todd S. Horowitz1,2, and Jeremy M. Wolfe1,21Brigham and Women’s Hospital and 2Visual Attention Lab, Harvard Medical SchoolAbstractExperiments have shown that people can rapidly determine if categories such as “animal” or “beach” are present in scenes that are presented for only a few milliseconds. Typically, observers in these experiments report on one prespecified category. For the first time, we show that observers can rapidly extract information about multiple categories. Moreover, we demonstrate task-dependent interactions between accumulating information about different categories in a scene. This interaction can be constructive or destructive, depending on whether the presence of one category can be taken as evidence for or against the presence of the other.Keywordsnonselective processing, attention, scenes, gist, interactionsReceived 9/10/10; Revision accepted 12/23/10Research Article Psychological Science OnlineFirst, published on May 9, 2011 as doi:10.1177/0956797611407930 at Harvard Libraries on May 22, 2011pss.sagepub.comDownloaded from2 Evans et al. potential filters may be large, maybe they cannot all be active simultaneously. Indeed, one might imagine that trying to extract more than one nonselective property from the same brief image might lead to destructive interference between fil-ters, such that selective attention to one property is required in order to disambiguate the information.Can two or more properties be extracted in a single nonselec-tive step? Suppose that two questions are asked about the same image: Is there an animal present? Is this a beach scene? It seems possible that the feed-forward sweep of information could provide answers for both of these questions. Alternatively, the system might need to be configured in advance to direct the information to one or the other classifier. Of course, one can also imagine intermediate states between these extremes. We conducted a series of


Download Psychological Science
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Psychological Science and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Psychological Science 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?