Feature integration theory revisited: Dissociating feature detection and attentional guidance in visual search.

Chan, Louis K. H.; Hayward, William G.

doi:10.1037/0096-1523.35.1.119

Cited by 43 publications

(101 citation statements)

References 59 publications

(99 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is indeed what the present results show, as do those of Theeuwes et al (2006) and those of Mortier et al (2005). Furthermore, our account is in line with the recent findings of Chan and Hayward (2009). They found a crossdimensional cost in a detection task but not in a localization task or in a compound task.…”

Section: Discussionsupporting

confidence: 82%

Word cues affect detection but not localization responses

Mortier

Zoest²,

Meeter³

et al. 2010

Attention, Perception, & Psychophysics

View full text Add to dashboard Cite

Many theories assume that preknowledge of an upcoming target helps visual selection. In those theories, a top-down set can alter the salience of the target, such that attention can be deployed to the target more efficiently and responses are faster. Evidence for this account stems from visual search studies in which the identity of the upcoming target is cued in advance. In five experiments, we show that top-down knowledge affects the speed with which a singleton target can be detected but not the speed with which it can be localized. Furthermore, we show that these results are independent of the mode of responding (manual or saccadic) and are not due to a ceiling effect. Our results suggest that in singleton search, top-down information does not affect visual selection but most likely does affect response selection. We argue that such an effect is found only when information from different dimensions needs to be integrated to generate a response and that this is the case in singleton detection tasks but not in other singleton search tasks.

show abstract

Section: Discussionsupporting

confidence: 82%

Word cues affect detection but not localization responses

Mortier

Zoest²,

Meeter³

et al. 2010

Attention, Perception, & Psychophysics

View full text Add to dashboard Cite

show abstract

“…In contrast with previous studies that employed this method (Nozawa, Reuter-Lorenz, & Hughes, 1994;Patching & Quinlan, 2004;Townsend & Nozawa, 1995), in the present experiments, we investigated how dimensional feature contrast signals are processed in visual search for feature singletons, according to saliency summation (e.g., Wolfe, 1994) or alternative models (e.g., serial checking, Chan & Hayward, 2009;Treisman & Gelade, 1980). Second, instead of requiring observers to make a simple manual response to the onset of one stimulus presented at a fixed location (Patching & Quinlan, 2004;Townsend & Nozawa, 1995) or saccade to a target stimulus presented at one of two possible locations (Nozawa et al, 1994), in the present study, the redundant-signals paradigm was implemented in a visual-search task in which a detection response had to be given if one (target) item differed from the surrounding (distractor) items in one or two dimensions (similar to Krummenacher et al, 2001, and Turatto et al, 2004.…”

Section: Methodsmentioning

confidence: 97%

“…Such a double-check strategy would result in increased RTs on single-target trials, provided that it takes longer, statistically, to determine the absence of a signal in a channel as compared with its presence, which is the norm in the relevant pop-out search studies (e.g., Found & Müller, 1996;Treisman & Gormican, 1988; see also Chun & Wolfe, 1996). Indeed, FIT (Treisman & Gelade, 1980;Treisman & Gormican, 1988; see also Chan & Hayward, 2009) assumes detection of singletons to involve a serial process of checking for activity in dimensional modules. Consequently, serial exhaustive search of dimensional modules has to be considered as an alternative to saliency summation models.…”

Section: Overview Of the Present Studymentioning

confidence: 99%

“…According to FIT, detection responses may be based directly on the activity in "dimensional modules" (which may be thought of as summing feature contrast signals within a given dimension; see Krummenacher et al, 2002); that is, a detection response is triggered as soon as the activity of such a module exceeds threshold. Note that Treisman's (1988) original assumption was that dimensional modules have to be checked for the presence of an above-threshold signal in a serial fashion (a possibility that will be of relevance in the section below; see Chan & Hayward, 2009, for a revival of this proposal). However, it is also conceivable that dimensional modules are processed in parallel; that is, a detection response may be triggered as soon as activity in one of the dimensional modules reaches threshold, which reduces to a classical race model (see Figure 1A).…”

Section: The Redundant-signals Paradigm In Visual Search For Singletomentioning

confidence: 99%

See 1 more Smart Citation

The detection of feature singletons defined in two dimensions is based on salience summation, rather than on serial exhaustive or interactive race architectures

Zehetleitner

Krummenacher

Müller

2009

Attention, Perception & Psychophysics

View full text Add to dashboard Cite

A common assumption made by most current models of visual search is saliency summation: Salience is computed in parallel in separate visual dimensions and is then summed onto a master salience map that guides the allocation of focal attention (see, e.g., Found & Müller, 1996;Gao, Mahadevan, & Vasconcelos, 2008;Itti & Koch, 2000;Koch & Ullman, 1985;Müller, Heller, & Ziegler, 1995;Wolfe, 1994;Wolfe, Cave, & Franzel, 1989). However, although it is by now a standard assumption, the empirical evidence for saliency summation is rather sparse (Krummenacher, Müller, & Heller, 2001Nothdurft, 2000;Poirier, Gosselin, & Arguin, 2008), and there are alternative (serial, parallel-independent, or parallel-interactive) processing architectures with various stopping rules (selfterminating or exhaustive search). The aim of the present study was to strengthen the support for the assumption of saliency summation, vis-à-vis the alternative accounts. Saliency Summation ModelsIn saliency summation models (see, e.g., Itti & Koch, 2000;Koch & Ullman, 1985;Treisman & Gelade, 1980;Wolfe, 1994), feature contrast is a measure of how different a specific location in the field is relative to its surrounding locations with regard to a particular feature. For example, with a red vertical bar surrounded by green vertical bars, feature contrast is high for "red" and low for "vertical." According to saliency summation models, feature contrast signals are pooled into dimension-specific maps (for an overview of dimensions in visual search, see Wolfe, 1998, andWolfe &Horowitz, 2004) and are then summed into a supradimensional saliency or master map of the visual field. Activation at any location on the master map signals the presence of local feature differences, without providing information about the critical dimensions or features that give rise to these differences. Ludwig-Maximilians-Universität München, Munich, GermanyInfluential models of visual search assume that dimension-specific feature contrast signals are summed into a master saliency map in a coactive fashion. The main source of evidence for coactivation models, and against parallel race models, is violations of the race model inequality (RMI;Miller, 1982) by redundantly defined singleton feature targets. However, RMI violations do not rule out serial exhaustive (Townsend & Nozawa, 1997) or interactive race (Mordkoff & Yantis, 1991) architectures. These alternatives were tested in two experiments. In Experiment 1, we used a double-factorial design with singleton targets defined in two dimensions and at two levels of intensity, to distinguish between serial versus parallel models and self-terminating versus exhaustive stopping rules. In Experiment 2, we manipulated contingency benefits that are expected to affect the magnitude of redundancy gains and/or RMI violations on the assumption of an interactive race. The results of both experiments revealed redundancy gains as well as violations of the RMI, but the data pattern excluded serial-exhaustive and interactive race models as possible ...

show abstract

“…Previous studies have demonstrated that in visual search and change detection experiments, identification is dependent on localization (Treisman & Gelade, 1980;Treisman & Gormican, 1988;Dukewich & Klein, 2009;Ghorashi et al, 2010;Busch et al, 2010a;Schneider & Wascher, 2013;Chan & Hayward, 2009;Watanabe, 2003;Wolfe et al, 2006;Agostinelli et al, 1986;Becker et al, 2000;Hughes et al, 2012). Thus, Experiment 2 investigated whether change detection without localization also implies a failure to identify the changing feature.…”

Section: Discussionmentioning

confidence: 99%

Change detection on a hunch: Pre-attentive vision allows “sensing” of unique feature changes

Ball

Busch

2015

Atten Percept Psychophys

View full text Add to dashboard Cite

Studies on change detection and change blindness have investigated the nature of visual representations by testing the conditions under which observers are able to detect when an object in a complex scene changes from one moment to the next. Several authors have proposed that change detection can occur without identification of the changing object, but the perceptual processes underlying this phenomenon are currently unknown. We hypothesized that change detection without localization or identification occurs when the change happens outside the focus of attention. Such changes would usually go entirely unnoticed, unless the change brings about a modification of one of the feature maps representing the scene. Thus, the appearance or disappearance of a unique feature might be registered even in the absence of focused attention and without feature binding, allowing for change detection, but not localization or identification. We tested this hypothesis in three experiments, in which changes either involved colors that were already present elsewhere in the display or entirely unique colors. Observers detected whether any change had occurred and then localized or identified the change. Change detection without localization occurred Moreover, change detection without localization for unique feature changes was independent of the number of objects in the display and independent of change identification. These findings suggest that pre-attentive registration of a change on a feature map can give rise to a conscious experience even when feature binding has failed: that something has changed without knowing what or where. Keywords

show abstract

Feature integration theory revisited: Dissociating feature detection and attentional guidance in visual search.

Cited by 43 publications

References 59 publications

Word cues affect detection but not localization responses

Word cues affect detection but not localization responses

The detection of feature singletons defined in two dimensions is based on salience summation, rather than on serial exhaustive or interactive race architectures

Change detection on a hunch: Pre-attentive vision allows “sensing” of unique feature changes

Contact Info

Product

Resources

About