Attentional Highbeams
 
 
Here are some demonstrations of the various conditions discussed in the following paper:
Flombaum, J. I., Scholl, B. J., & Pylyshyn, Z. W. (2008). Attentional resources in visual tracking through occlusion: The high-beams effect. Cognition, 107(3), 904-931.
These demonstrations are provided as Quicktime movies, which can be downloaded or viewed directly in most web-browsers. (To download a free Quicktime player, go here.) These movies are a bit large and choppy, but they should be sufficient to illustrate the basic conditions. As highly compressed versions of the original stimuli, these movies may not preserve the precise spatial and temporal characteristics of the originals.  
 
This project investigated the on-line resources that support the ability to keep track of objects through periods of occlusion. We explored how attention is distributed when featurally identical objects become occluded during multiple object tracking. During tracking, observers had to detect small probes that appeared sporadically on targets, distractors, occluders, or empty space. Probe detection rates for these categories were taken as indexes of the distribution of attention throughout the display and revealed three key effects, as described below.  
 
Sample Animation from Experiment 1 (560 KB)
Sample Animation from Experiment 1 with 'Pop-up' Notes (640 KB)
 
These movies depict the probe types used in Experiment 1. Observers tracked 3 out of 6 moving discs, and had to detect brief gray probes that could occur in several potential locations. (The 'pop-up' movie highlights these different probe types, which are explained in detail in the manuscript.) There were three key results: (1) Probes on distractors were detected worse than when they occurred nearby in empty space, indicating that distractors in MOT are still tracked via some type of inhibitory processing (as first noted by Pylyshyn, 2006). (2) Probe detection was better on occluders when they were currently occluding objects -- indicating that object-based attention can be directed even to invisible objects. (3) Most important and surprising probe detection for both targets and distractors was always better when they were occluded, compared to when they were visible. This new attentional high-beams effect indicates that the ability to track through occlusion, though seemingly effortless, in fact requires the active allocation of special attentional resources: it seems that the maintenance of persisting object representations through occlusion poses some special challenge, for which an added burst of attention is allocated, much like the way that one turns on the high-beams on a car to illuminate the road when it is especially hard to see. This effect was especially striking given (1) that tracking through occlusion seems largely effortless (relative to tracking without occlusion), and (2) that it doesn't feel like distractors are tracked at all, much less with special intensity during occlusion.  
 
Sample Animation from Experiment 2 (2.8 MB)
Sample Animation from Experiment 2 with 'Pop-up' Notes (2.8 MB)
 
This experiment was identical to Experiment 1, except that we used stationary objects and moving occluders (rather moving objects and static occluders). This manipulation tested whether the the main effects from Experiment 1 arose from the demands of tracking the moving targets, or whether they could somehow be explained in terms of lower-level differences in probe visibility across conditions. Though these visual differences were the same in this experiment, the relevant effects disappeared: we did not find amplified attention to occluders with objects behind them, relative to either (1) occluders with nothing behind them, or (2) unoccluded objects. We conclude that the 'attentional high-beams' effect reflects the special demands of tracking objects through occlusion.  
 
Sample Animation from Experiment 3 (2.4 MB)
Sample Animation from Experiment 3 with 'Pop-up' Notes (2.7 MB)
 
This experiment replicated the three primary effects from Experiment 1, but now using rather different displays. Here the background, the objects, and the interiors of the occluders were all defined by visual noise, with each pixel set randomly to either black or white. Thus the objects did not exist in any static frame, but were simply defined by their motion-defined coherence. (Of course, it is interesting in the first place that subjects can still track under such conditions!) All three effects from Experiment 1 replicated here, indicating (1) that they were not dependent on any subtle differences in the objects' contours (since those contours did not exist here!); and (2) that these effects generalize across both large superficial display differences and different degrees of attentional load. Again, the 'pop-up' version highlights the different probe types.  
 
Sample Animation from Experiment 4 (2.2 MB)  
This experiment was identical to Experiment 3, except that the occluders weren't functional: instead, the motion-defined objects just passed right over their contours. Here the attentional high-beams effect disappeared, demonstrating the this effect is dependent on the existence of occlusion, per se.