290 likes | 671 Views
Visual attention reveals changing color in moving objects James E. Hoffman and Scott McLean University of Delaware. Change Blindness. Alternating pictures differing in one feature or object Usually accompanied by a mask or irrelevant transient, that results in “change blindness”.
E N D
Visual attention reveals changing color in moving objectsJames E. Hoffman and Scott McLeanUniversity of Delaware
Change Blindness Alternating pictures differing in one feature or object Usually accompanied by a mask or irrelevant transient, that results in “change blindness”
Attention and Change Detection The transient in the second presentation automatically captures attention, leading to awareness of the changing feature. The “mudsplashes” in the first presentation provide additional transients that mask those produced by the changing lane marker. If attention is directed to the location of the change, awareness of the change is restored. Note that this experiment also reveals the limited-capacity nature of visual attention. Apparently we can only attend to a few objects in the picture at any given time.
Motion-induced change blindness Suchow and Alvarez (2011) recently introduced a striking form of change blindness that may require a different mechanism to explain a failure of awareness for change. In their illusion, all the objects are changing instead of just one and the illusion “persists even though the observer attends to the objects” and “knows that they are changing” (abstract) They called this effect “motion silencing”
Motion Silencing of Color Change The same effect occurs for changes in shape, lightness, and size
What Causes Motion Silencing? • David Burr, in a commentary accompanying the original article, suggested that two factors are responsible for Motion Silencing: • Motion coherence • When a crowd of objects moves coherently together, perceiving the individual objects may be difficult (Poljac et al., 2012). Similar to “global precedence” effect. • Crowding • Close objects may not be individuated. In this case, the observer may not be able to assign successive colors to a single object. In contrast, stationary dots will still stimulate retinotopic change detectorsthat can signal the presence of a change (Turi and Burr, 2013).
The role of attention in motion silencing We suggest that in motion silencing, the color change is masked by transients associated with motion and that locating the color change requires attention, just like the mud splash experiment. Coherent motion and crowding are not the causes of the motion silencing effect; they simply modulate its strength through their effect on the real causal mechanism which is visual attention. Coherent motion and crowding both increase the difficulty of paying attention to individual objects but the illusion should still occur when motion coherence and crowding are eliminated, as long as the display exceeds the observer’s ability to attend to all of the objects in parallel.
Current Study In the current study, we evaluated the effects of attention by asking observers to track 1, 2, or 3 target objects embedded in a display of distractors (MOT). Detection of color change should be much better for changes occurring on targets compared to distractors and should be closely related to tracking accuracy . In contrast, performance with static displays should be high and independent of set size. We also attempted to determine whether silencing can be obtained with incoherent motion and uncrowded displays which would show that these variables are not crucial ingredients of the motion silencing effect.
Trial Events Tracking Trials No Color Change Click on Target Dots Color Change Trials Motion (2 ½ sec.) Static Display Cue Display Set Size =3 Color Change occurs on a Target or a Distractor Moving Stationary
Results Replication of the motion silencing effect Attend all condition Observers missed approximately 1% of the changes with static displays while they missed 36% of the changes with moving displays. This is essentially a replication of the motion silencing effect for our paradigm.
Results: Tracking Accuracy Observers were limited in their ability to accurately track more than a few objects
Change Detection Attend All
Static Control Percent Correct set size Change Detection was at ceiling in the stationary condition and was unaffected by set size or attention Change Detection was mediated by retinotopic change detectors
Does change detection on distractors depend on their distance from a target? But it does depend on distance from fixation No .. Close: 1.42deg to 1.61 dva Medium: 1.63 to 2.02 dva Far: > 2.03 dva
Summary Like previous studies on change blindness, detection of change in motion silencing was heavily dependent on whether or not the change was attended. In the next experiment, we show that motion silencing still occurs when displays are “uncrowded” and still depends on attention being allocated to the changing object
Expt. 2: Silencing with “Uncrowded” Displays + “Sparse” Displays containing only 10 objects Observers tracked two objects or attended to all of them Objects appeared as either stationary or moving
Silencing with “Uncrowded” Displays + Minimum inter-object spacing was large enough to eliminate crowding for the most peripheral objects, i.e., there is no crowding in these displays. Do we still get motion silencing?
Results Tracking accuracy Moving: 94.7% Stationary: 98.1% Uncrowded Corresponding Data from Expt. 1.
Conclusions • Neither crowding nor motion coherence are necessary conditions for motion silencing • Motion silencing may be due to masking of change signals by object motion • Attention to individual moving objects isolates them from these masking signals allowing the change signal to emerge • Effects of crowding and coherence may be seen as factors that modulate the difficulty of attending to individual objects • Color change detection appears be a useful tool for studying the role of visual attention in MOT
Does attention affect motion silencing? Suchow and Alvarez suggested that it might: “Conceivably, the brief exposure afforded by a fast-moving object could be lengthened by rapidly shifting the focus of attention over a moving window [10, 11]. Here, tight spacing precluded isolation, and explicit instructions to pay attention to the whole set discouraged any shifts of attention. Manipulating the number of dots or asking observers to attend to only a few of them might reveal the role of attention in silencing.” (p. 141) But Turi and Burr suggest that ‘‘motion silencing seems to resist attention to individual dots.” (p.1)