All eyes on relevance: strategic allocation of attention as a result of feature-based task demands in multiple object tracking

Abstract

Multiple object tracking (MOT) plays a fundamental role in processing and interpreting dynamic environments. Regarding the type of information utilized by the observer, recent studies reported evidence for the use of object features in an automatic, low-level manner. By introducing a novel paradigm that allowed us to combine tracking with a noninterfering top-down task, we tested whether a voluntary component can regulate the deployment of attention to taskrelevant features in a selective manner. In four experiments we found conclusive evidence for a task-driven selection mechanism that guides attention during tracking: The observers were able to ignore or prioritize distinct objects. They marked the distinct (cued) object (target/distractor) more or less often than other objects of the same type (targets /distractors)-but only when they had received an identification task that required them to actively process object features (cues) during tracking. These effects are discussed with regard to existing theoretical approaches to attentive tracking, gaze-cue usability as well as attentional readiness, a term that originally stems from research on attention capture and visual search. Our findings indicate that existing theories of MOT need to be adjusted to allow for flexible top-down, voluntary processing during tracking.

Goal-directed unequal attention allocation during multiple object tracking

“…Results from all experiments revealed some evidence for unequal attention allocation according to strategic top-down control. This is in line with the existing literature documenting top-down, goal-driven attention allocation in MOT (Brockhoff & Huff, 2016) and visual search (Jiang et al, 2015; Navalpakkam et al, 2010; Shomstein & Johnson, 2013). Such findings replicate research demonstrating unequal attention allocation during MOT in response to instructions (Cohen, Pinto, Howe, & Horowitz, 2011; Yantis, 1992), further supporting the efficacy of using goal-directed instructions to manipulate participants’ attention allocation (Bonnel & Miller, 1994; Fitsoul, 2016).…”

“…This shows that participants modified their tracking strategy in a goal-directed manner. Brockhoff and Huff (2016) combined a typical MOT task with a non-interfering top-down identification task. Participants were instructed to identify the behavior of dynamic cartoon eyes.…”

Attentional modulation of hierarchical ensemble coding for the identities of moving faces.

“…The findings of the present study may also have implications for our understanding of the attentional mechanisms involved in the MOT paradigm. Several studies have suggested that observers distributed visual attention in a hierarchical manner over the entire display in the MOT task (Brockhoff & Huff, 2016; Doran & Hoffman, 2010; Drew, McCollough, Horowitz, & Vogel, 2009) because the distractors were being processed while the observers were tracking the targets (Alvarez & Oliva, 2008; Brockhoff & Huff, 2016; Doran & Hoffman, 2010; Drew et al, 2009; Meyerhoff, Papenmeier, Jahn, & Huff, 2015, 2016; Tran & Hoffman, 2016). Our hierarchical representation finding is consistent with this idea, which should be considered in future studies of MOT.…”

Hue distinctiveness overrides category in determining performance in multiple object tracking

“…When tracking objects with an identity that is irrelevant to the task (the identity is never tested), the question is no longer whether the identity information is processed during tracking. Rather, researchers are more concerned about what kinds of information can be coded and why (Brockhoff & Huff, 2016;Erlikhman et al, 2013;Feria, 2012;Liu et al, 2012;). According to previous studies, both voluntary and involuntary identity processing can be observed in MOT (Erlikhman et al, 2013;Liu et al, 2012;Makovski & Jiang, 2009a, b).…”

“…When color category is involved, the case becomes more complex in MOT. First, MOT is a goal-directed task where participants can voluntarily adopt strategies to aid tracking (Brockhoff & Huff, 2016;Feria, 2012). Thus, they may use distinct color labels to name objects when the colors are from different categories and store them in verbal working memory.…”