Target-distractor Discriminability Research Articles

On the relationship between target-distractor discriminability and search efficiency: the case of color.

On the relationship between target-distractor discriminability and search efficiency: the case of color.

Effect of L-theanine on selective attention in a traffic-related reaction task in sleep-deprived young adults: a double-blind placebo-controlled, crossover study

ABSTRACT Background L-theanine is a non-protein-forming amino acid found in tea. Limited evidence suggests that it improves selective attention. Sleep deprivation impairs attention and psychomotor reactions, affecting automobile driving. We aimed to determine whether L-theanine improves neurobehavioral measures of visual attention in acutely sleep-deprived healthy adults in a traffic-scene-based attention task. Methods In a double-blind, placebo-controlled, counterbalanced, two-way crossover study, we compared the effects of a 200-mg dose of L-theanine with a placebo (150 ml of distilled water) on a computerised, traffic-scene-based visual recognition reaction task in 24 healthy volunteers (age 20–25 years; 13 males) sleep-deprived overnight. The participants made speeded button-presses to imminent accident scenes (i.e. hits), while ignoring safe scenes. They were tested pre-dose and 45 min post-dose, each treatment administered one week apart. Results Hit rates were more than 90% in all sessions, and were similar in two treatments, pre- vs post-dose. L-theanine significantly reduced false alarms (i.e. responses to safe scenes) (p = 0.014) and increased A′ (i.e. target-distractor discriminability) (p = 0.009), whereas placebo did not (p > 0.05). L-theanine reduced hit reaction time by 38.65 ms (p = 0.007), and placebo by 19.08 ms (p = 0.016), however reaction time changes from baseline were not significantly different between treatments (p > 0.05). Conclusions L-theanine in high doses appears to improve selective visual attention by concurrently improving information processing speed and target-distractor discriminability in acutely sleep-deprived individuals. This is consistent with previous functional neuroimaging findings, where L-theanine suppressed distractor-processing and default-mode-network activity in visual selective attention tasks.

Where to display vital information? ERP evidence for background changes

Critical decision systems require expeditious and accurate responses to the displayed information. In addition to content, location and background are equally important. Several visual search studies have pointed out the differences and compared modes of top-down attention allocation: distractor suppression and attentional capture. Previous studies have used color (mainly) and shape as a feature but have overlooked luminance as a feature for studying underlying attention mechanisms. The present study attempts to bridge this gap. In this study, participants performed a target-distractor discrimination task by identifying a randomly appearing target from the pool of distractors based on defined luminance levels. Background change was noticed by manipulating the task such as making visible quadrant boundaries over the screen. The preliminary evidence suggested that displaying information at the top-left of the screen had higher percentage accuracy; whereas, response time (RT) remained unaffected. Improvement in RT and percentage accuracy was observed with task manipulation. Event-related potential (ERP) analysis revealed elicited Distractor Positivity (PD), providing evidence for the distractor suppression hypothesis. Further, differences emerged in the topographic plot of N2pc and PD. In sum, the result contributes to classic debate of capture vs. suppression and provides a crucial connection between display design and electrophysiological indices, emphasizing locations and background as equally important factors.

Target templates in low target-distractor discriminability visual search have higher resolution, but the advantage they provide is short-lived.

When you search repeatedly for a set of items among very similar distractors, does that make you more efficient in locating the targets? To address this, we had observers search for two categories of targets among the same set of distractors across trials. Visual and conceptual similarity of the stimuli were validated with a multidimensional scaling analysis, and separately using a deep neural network model. After a few blocks of visual search trials, the distractor set was replaced. In three experiments, we manipulated the level of discriminability between the targets and distractors before and after the distractors were replaced. Our results suggest that in the presence of repeated distractors, observers generally become more efficient. However, the difficulty of the search task does impact how efficient people are when the distractor set is replaced. Specifically, when the training is easy, people are more impaired in a difficult transfer test. We attribute this effect to the precision of the target template generated during training. In particular, a coarse target template is created when the target and distractors are easy to discriminate. These coarse target templates do not transfer well in a context with new distractors. This suggests that learning with more distinct targets and distractors can result in lower performance when context changes, but observers recover from this effect quickly (within a block of search trials).Supplementary InformationThe online version contains supplementary material available at 10.3758/s13414-020-02213-w.

What limits search for conjunctions of simple visual features?

Despite of decades of research, we still do not know for sure the roles of internal noise, attention, and crowding in search for conjunctions of simple visual features. In this study, we tried several modifications to the classic design of conjunction-search experiments. In order to match exactly the proportions of simple features, two different targets were presented in target-present trials-vertical red and horizontal blue bars among vertical blue and horizontal red distractors. Both the length of the bars and the number of objects in a display were varied. Positions of objects were selected for minimal crowding effects. Exposure duration was 60 ms, and proportion correct was used as the measure of performance. For conjunction search, the results rejected the unlimited-capacity model and were consistent with limited-capacity attentional processing, and with the Naka-Rushton transform of the target-distractor difference. Qualitatively the same results were obtained when bar length was fixed, and fine orientation difference was used to manipulate target-distractor discriminability. An experiment of feature (orientation) search produced results close to the unlimited-capacity model.

Measuring visual search and distraction in immersive virtual reality.

An important issue of psychological research is how experiments conducted in the laboratory or theories based on such experiments relate to human performance in daily life. Immersive virtual reality (VR) allows control over stimuli and conditions at increased ecological validity. The goal of the present study was to accomplish a transfer of traditional paradigms that assess attention and distraction to immersive VR. To further increase ecological validity we explored attentional effects with daily objects as stimuli instead of simple letters. Participants searched for a target among distractors on the countertop of a virtual kitchen. Target–distractor discriminability was varied and the displays were accompanied by a peripheral flanker that was congruent or incongruent to the target. Reaction time was slower when target–distractor discriminability was low and when flankers were incongruent. The results were replicated in a second experiment in which stimuli were presented on a computer screen in two dimensions. The study demonstrates the successful translation of traditional paradigms and manipulations into immersive VR and lays a foundation for future research on attention and distraction in VR. Further, we provide an outline for future studies that should use features of VR that are not available in traditional laboratory research.

Camouflaging moving objects: crypsis and masquerade.

Motion is generally assumed to "break" camouflage. However, although camouflage cannot conceal a group of moving animals, it may impair a predator's ability to single one out for attack, even if that discrimination is not based on a color difference. Here, we use a computer-based task in which humans had to detect the odd one out among moving objects, with "oddity" based on shape. All objects were either patterned or plain, and either matched the background or not. We show that there are advantages of matching both group-mates and the background. However, when patterned objects are on a plain background (i.e., no background matching), the advantage of being among similarly patterned distractors is only realized when the group size is larger (10 compared to 5). In a second experiment, we present a paradigm for testing how coloration interferes with target-distractor discrimination, based on an adaptive staircase procedure for establishing the threshold. We show that when the predator only has a short time for decision-making, displaying a similar pattern to the distractors and the background affords protection even when the difference in shape between target and distractors is large. We conclude that, even though motion breaks camouflage, being camouflaged could help group-living animals reduce the risk of being singled out for attack by predators.

Unchanging Capacity

Complex workspaces often require operators to divide attention between information within the visual periphery and a visual central task. For an air traffic controller, for example, monitoring complex displays while also watching for potential hazards is essential for avoiding aircraft collisions. In such environments, fast and accurate detection of peripheral events may be critical for safe performance. Presenting targets redundantly offers a potential way of speeding up target detection (Little, Eidels, Fific, & Wang, 2015; Townsend & Eidels, 2011). It remains unclear, however, whether redundant-target processing remains efficient with a concurrent central task. A series of experiments examined the effects of dual-tasking on peripheral redundant-target processing, either between- (Experiments 1a & 1b) or within-participants (Experiment 4). Furthermore, Experiments 2 and 3 manipulated target-distractor discriminability and distractor presence to examine the effects of target salience on dual-task processing efficiency. One hundred and one undergraduate students ( N =20 in each of Experiments 1a to 3; N = 21 in Experiment 4) performed a redundant-target task either by itself (Experiment 1a) or whilst performing a manual tracking task (Experiments 1b-4). The tracking task required participants to maneuver a joystick using both hands to align a cursor with a moving red target. The detection task required participants to press a joystick button bimanually whenever a target appeared at a location in the peripheral visual field. Experiments 1a, 1b, and Experiment 4 employed “T” as the target item and “L” as distractor items that appeared randomly rotated in 90° steps. In Experiment 2, target salience was increased by employing “X” as the target item and “O” for the distractor items. Experiment 3 tested peripheral target processing in the absence of distractors; hence, only the target item “T” was employed. Processing efficiency in the target-detection task was calculated using measures of resiliency (Little et al., 2015) or workload capacity (Townsend & Eidels, 2011). In all five experiments, processing of redundant targets was less efficient than predicted by a standard parallel race model (Raab, 1962; Townsend & Eidels, 2011). Surprisingly, processing efficiency differed negligibly between the single and dual-task conditions. Capacity may be protected from task-load effects due to separate information-processing resource pools within the central and peripheral visual fields (Wickens, 2002). Neither increasing the discriminability between targets and distractors, nor removing distractors entirely, had any effect on redundant-target processing efficiency. Results suggest target processing in the visual periphery is capacity-limited, but that processing efficiency is robust against changes to concurrent task load or target salience.

Distinct load effects by set-size and target-distractor similarity in visual search

Using the simultaneous-sequential presentation search paradigm, we investigated how different types of load manipulation would tap into attentional resources and the concrete mechanism by which search would be performed. Search display perceptual load was manipulated by set-size and target-distractor similarity. The benefit of sequential presentation was larger when the load was increased by number of search items than when target-distractor discrimination was made more demanding. Considering that the load effect could be result from both perceptual load and statistical decision noise, the current results are explained by suggesting that set-size will determine whether the search process will be serial or parallel, regardless of perceptual difficulty. Factors that can set the limit of attentional resources are also discussed.

Task-irrelevant stimulus-reward association induces value-driven attentional capture.

Rewards affect the deployment of visual attention in various situations. Evidence suggests that the stimulus associated with reward involuntary captures attention (value-driven attentional capture; VDAC). Recent studies report VDAC even when the reward-associated feature does not define the target (i.e., task-irrelevant). However, these studies did not conduct the test phase without reward, thus the effect may be qualitatively different from those in the previous studies. In the current study, we tested if task-irrelevant features induce VDAC even in the test phase with no reward. We used a flanker task during reward learning to create color-reward associations (training phase), and then tested the effect of color during visual search (test phase). Reward learning with no spatial uncertainty in the flanker task induced VDAC, even when reward signaling color was associated with both target and distractor (Experiments 1 and 2). In Experiment 3, a significant VDAC with a color for all letters indicated that target-distractor discrimination is not necessary for VDAC. Finally, a significant VDAC (Experiment 4) with color rectangular frames around the letters indicated binding reward-associated features to task-relevant letters is not necessary for VDAC. All these effects were obtained in the test phase without reward, thus VDAC in the current study is comparable to previous studies using target-defining features. These findings indicate that task-relevance is not a necessary condition for VDAC from reward-associated features, suggesting that reward-associated learning in VDAC is more indirect.

The capacity limitations of orientation summary statistics.

The simultaneous-sequential method was used to test the processing capacity of establishing mean orientation summaries. Four clusters of oriented Gabor patches were presented in the peripheral visual field. One of the clusters had a mean orientation that was tilted either left or right, whereas the mean orientations of the other three clusters were roughly vertical. All four clusters were presented at the same time in the simultaneous condition, whereas the clusters appeared in temporal subsets of two in the sequential condition. Performance was lower when the means of all four clusters had to be processed concurrently than when only two had to be processed in the same amount of time. The advantage for establishing fewer summaries at a given time indicates that the processing of mean orientation engages limited-capacity processes (Exp. 1). This limitation cannot be attributed to crowding, low target-distractor discriminability, or a limited-capacity comparison process (Exps. 2 and 3). In contrast to the limitations of establishing multiple summary representations, establishing a single summary representation unfolds without interference (Exp. 4). When interpreted in the context of recent work on the capacity of summary statistics, these findings encourage a reevaluation of the view that early visual perception consists of creating summary statistic representations that unfold independently across multiple areas of the visual field.

Can attenuation of attentional blink also evoke removal of repetition blindness?

A recent study showed that attentional blink (AB), which has been considered to reflect the capacity limitation of visual temporal attention, can be attenuated after a short period of the color-salient training, in which the second target (T2) within the AB period is given a salient color (Choi et al., 2012). The current study explored whether the effect of the color-salient training could be transferred to another phenomenon. In addition to AB, repetition blindness (RB) was employed, which is phenomenologically similar to, but fundamentally different from AB. After completion of the color-salient training with a nonrepeated T2 (corresponding to AB), RB was still observed, whereas AB was completely removed. However, the color-salient training with a repeated T2 (similar to RB) induced not only a significant reduction of RB but also an attenuation of AB. This result provides further evidence for dissociation between AB and RB. In addition, it implies that the color-salient training improves the attentional control mechanism related to target-distractor discrimination rather than to the perceptual system.

The role of search difficulty in intertrial feature priming

Previous research has shown that intertrial repetition of target and distractors task-relevant properties speeds visual search performance, an effect known as priming of pop-out (PoP). Recent accounts suggest that such priming results, at least in part, from a mechanism that speeds post-selectional, response-related processes, the marker of which is an interaction between repetition of the target and distractor features and repetition of the response from the previous trial. However, this response-based component of inter-trial priming has been elusive, and it remains unclear what its boundary conditions might be. In addition, what information is represented in the episodic memory traces that underlie the response-based component has not yet been characterized. Here, we show that the response-based component of feature priming reflects an episodic memory retrieval mechanism that is not mandatory or automatic but may be described as a heuristic that subjects sometimes use, in particular when the overall difficulty of the search task is high. In addition, we show that the conjunction of the target and distractor features forms the context that is reactivated during episodic retrieval. Finally, we show that target–distractor discriminability is an important modulator of the selection-based component. The findings are discussed within the framework of the dual-stage model of inter-trial priming ( Lamy, Yashar, & Ruderman, 2010).

Thresholds for the Detection of Changing Visual Features

Change blindness refers to the difficulty observers have in detecting otherwise obvious changes to visual stimuli, when these changes are masked in some way. Typically, change blindness is studied by using complex visual scenes and complex changes to these scenes. In the current study, we used a more controlled visual environment, presenting observers with a series of oriented, sinusoidal patterns (Gabors), one of which underwent a change during a blanking of the screen. Changes were made to different features (size, colour, spatial frequency, and speed) with the target-distractor discriminability varying. The detectability of these changes was quantified by calculating psychometric functions and thresholds for each individual observer. Thresholds for the detection of changing features were higher than those for non-changing features, but thresholds for both tasks show consistency across observers. Psychometric-function slopes show consistency across observers and change type only for non-changing targets. For changing targets, psychometric-function slopes show no obvious pattern across observers or change types. We suggest this reflects vSTM treating different features as abstract, interchangeable tokens, as alternative explanations (such as additional noise in vSTM) can be ruled out.

Surround Suppression Sharpens the Priority Map in the Lateral Intraparietal Area

In the visual world, stimuli compete with each other for allocation of the brain's limited processing resources. Computational models routinely invoke wide-ranging mutually suppressive interactions in spatial priority maps to implement active competition for attentional and saccadic allocation, but such suppressive interactions have not been physiologically described, and their existence is controversial. Much evidence implicates the lateral intraparietal area as a candidate priority map in the macaque (Macaca mulatta). Here, we demonstrate that the responses of neurons in the lateral intraparietal area (LIP) to a task-irrelevant distractor are strongly suppressed when the monkey plans saccades to locations outside their receptive fields. Suppression can be evoked both by flashed visual stimuli and by a memorized saccade plan. The suppressive surrounds of LIP neurons are spatially tuned and wide ranging. Increasing the monkey's motivation enhances target-distractor discriminability by enhancing both distractor suppression and the saccade goal representation; these changes are accompanied by correlated improvements in behavioral performance.

What, when and how of target detection in visual search

When and how do we decide whether the target is present in the scene or not? There are surprisingly few quantitative models that address this question. We propose a generative model of visual search. First, the observer obtains stimulus information (corrupted by additive white noise of variance σ) at each location and time instant. Second, information across scene locations is combined using optimal Bayesian inference to compute the instantaneous evidence for target presence in the scene (log likelihood ratio of target presence vs. absence at time t). Third, the observer accumulates evidence over time and decides ‘yes’ if the accumulated evidence exceeds criterion ζ+Δ, ‘no’ if the evidence falls below criterion −ζ+Δ, and otherwise continues to observe the scene. Δ is a criterion-shift which depends on the target frequency, and the reward for different responses. This model has only 2 free parameters (noise σ, criterion ζ), compared to many parameters in drift diffusion models. The model can explain several search phenomena including how accuracy and RT are affected by set-size, target-distractor discriminability, distractor heterogeneity, target frequency and reward. It can capture the shape of the RT distribution in target present and absent trials for different tasks like feature, conjunction and spatial configuration search. It explains that rare targets are missed (Wolfe et. al, 2005) because decreasing target frequency (e.g., from 50% to 2%) shifts the starting point of the decision process closer to ‘no’ than ‘yes’ criterion, leading to high miss error rates, and fast abandoning of search. Our model predicts that increasing penalties on miss errors will decrease these errors and increase RTs. We validated these predictions through psychophysics experiments with 4 human subjects. To summarize, we have proposed a generative model of visual search that with only 2 free parameters, can explain a wide range of search phenomena.

Reducing Satisfaction of Search Errors in Visual Search

Several occupations rely upon the ability to accurately and efficiently perform visual search. For example, radiologists must successfully identify abnormalities, and airport luggage screeners must recognize threatening items. By investigating various aspects of such searches, psychological research can reveal ways to improve search performance. We (Fleck, Samei, & Mitroff, in press) have recently focused on one specific influence that had previously been explored almost exclusively within the study of radiology: Satisfaction of Search (SOS), wherein the successful detection of one target can reduce detection of a second target in the same search array. To eliminate SOS errors, we must delineate the sources of the errors. Combined with our prior work, here we examine the specific roles of target heterogeneity and the decision-making component of target-distractor discrimination (i.e., how easy it is to determine whether a stimulus is a target or a distractor). In the current experiments, subjects searched arrays of line-drawn objects for targets of two different categories (tools and bottles) amongst several categories of distractor objects. On any given trial, there could be no targets, one target (either a tool or bottle), or two targets (both a tool and a bottle). The relative occurrence of these trial types varied across experiments. Compared to previous experiments that employed homogeneous targets and that required effortful evaluation to discriminate targets from distractors, the SOS effect was reduced here. That is, search accuracy for easily discriminable, heterogeneous objects was no worse for dual-target trials than for single-target trials. These results suggest that target heterogeneity and target-distractor discriminability may both play key roles in multiple target search accuracy.

The effect of feature discriminability on the intertrial inhibition of focused attention

Identifying the shape of a colour oddball is faster when the distractor colour is viewed in the preceding target-absent trial and slower when the target colour is previewed, an intertrial effect known as the distractor previewing effect (DPE). We tested the effect of feature discriminability on the DPE. In Experiment 1, we determined the interitem discriminability of two colour pairs and two shape pairs. In Experiments 2 and 3, we measured DPEs with these set of target–distractor discriminability pairs. Our results showed that when the defining features allow for efficient parallel search, the a priori degree of interitem discriminability did not modulate the DPE. The results suggest the DPE does not arise as a strictly bottom-up modulation of saliency of the search-relevant features but reflects an attentional bias aimed at preventing attention from revisiting recently rejected “search features”. The underlying mechanism of this attentional bias is discussed.

Do multiple object tracking and letter identification use the same visual attention resource?

Humans can track 3–5 independently moving objects among identical moving distractors (multiple object tracking). This has been taken as a demonstration of simultaneous parallel attention. Does this distributed attention facilitate letter identification, or do tracking and letter identification employ separate attentional resources? Eight participants tracked 2, 3, or 4 out of 8 disks for 5–10 seconds. Two trial types were randomly intermixed. On tracking probe trials, one disk turned white and participants made a 2AFC target-distractor discrimination. On letter probe trials, preand post-masked letters were flashed on one target and three distractors locations for 80 ms. Participants made a 13AFC decision about the letter presented on the target disk (possible letters: ABCDEHLNSTXYZ). Tracking accuracy was good: .92, .88, and .82 for 2, 3, and 4 targets, respectively. Letter accuracy was .69, .55 and .45. We used data from 2-target trials to predict letter accuracy for 3 and 4 targets using two models. The separate resource model assumes that not all tracked objects are attended but any letter flashed on an attended object is perfectly identified. Conversely, the unlimited unified resource model assumes that all tracked objects are attended but that letters flashed on attended objects are not perfectly identified. The separate model estimated 1.3 items simultaneously attended, predicting letter identification accuracy of .48 (3 targets) and .38 (4 targets), significantly underpredicting performance. The unified model estimated letter identification accuracy for letters on attended targets to be .71, predicting letter identification accuracy of .65 (3 targets) and .62. (4 targets), significantly overpredicting performance. The best explanation for the data is an (admittedly ad hoc) limited capacity unified resource model in which all tracked items are attended but the probability of identifying a letter on a tracked object declines as the number of tracked items increases.

Repetition advantage: Effects of inter-target and target-distractor discriminability

Repetition advantage: Effects of inter-target and target-distractor discriminability