Academic journal article Attention, Perception and Psychophysics

Direct versus Indirect Processing Changes the Influence of Color in Natural Scene Categorization

Academic journal article Attention, Perception and Psychophysics

Direct versus Indirect Processing Changes the Influence of Color in Natural Scene Categorization

Article excerpt

We examined whether participants would use a negative priming (NP) paradigm to categorize color and grayscale images of natural scenes that were presented peripherally and were ignored. We focused on (1) attentional resources allocated to natural scenes and (2) direct versus indirect processing of them. We set up low and high attention-load conditions, based on the set size of the searched stimuli in the prime display (one and five). Participants were required to detect and categorize the target objects in natural scenes in a central visual search task, ignoring peripheral natural images in both the prime and probe displays. The results showed that, irrespective of attention load, NP was observed for color scenes but not for grayscale scenes. We did not observe any effect of color information in central visual search, where participants responded directly to natural scenes. These results indicate that, in a situation in which participants indirectly process natural scenes, color information is critical to object categorization, but when the scenes are processed directly, color information does not contribute to categorization.

Natural scenes are readily processed by the visual system. In their seminal article, Thorpe, Fize, and Marlot (1996) found that go versus no-go natural scene categorization was processed rapidly and accurately. Subsequent studies concerning natural scene categorization have reported similar results (e.g., Bacon-Macé, Macé, Fabre- Thorpe, & Thorpe, 2005; VanRullen & Koch, 2003; VanRullen & Thorpe, 2001a, 2001b). Li, VanRullen, Koch, and Perona (2002) reported that, as compared with letters and color patterns, rapid natural scene categorization required very little attention and that higher level representations could be accessed in the near absence of attention. Also, Evans and Treisman (2005) showed that, whereas object categorization in natural scenes did not require attention, identification did.

We recently examined whether natural scenes were categorized with minimal attentional resources by using a negative priming (NP) paradigm (Otsuka & Kawaguchi, 2007; see also Lavie & Fox, 2000). Generally, in an NP experiment, participants are required to respond to a target while ignoring a distractor in each trial. What is important is the relationship between the distractor in trial n (i.e., the prime) and the target in trial n+1 (i.e., the probe). In the control (C) condition, no relation exists between the target and the distractor in the prime and probe trials. In the ignored repetition (IR) condition, the distractor ignored in the prime trial is presented as the target in the subsequent probe trial. In general, the reaction times (RTs) to the probe target in the IR condition are slower than those in the C condition. This delay is referred to as negative priming. The NP paradigm, used by Lavie and Fox, allowed us to measure the influence of the ignored images, which participants did not need to respond to. Our previous study compared natural scene categorization with letter discrimination and line drawing categorization. We set up low and high attention-load conditions, related to the size of the set of searched stimuli. In the low attention-load condition, one target and one distractor appeared in the prime display. In the high attention-load condition, one target was accompanied by four nontargets. On the prime display, all the distractors were incompatible with the response required to the current target, and all appeared randomly and equally either above or below the horizontal axis of the screen. In the C condition, the probe display contained a target and distractor that were different from those in the prime display. In the IR condition, a target on the probe display had the same identity as the distractor in the previous prime display. We predicted that, if people could categorize a target object in a natural scene with minimal attention, the NP effect should occur, irrespective of the attention load. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.