Local item density modulates adaptation of learned contextual cues

Abstract

In everyday scenes, searched-for targets do not appear in isolation, but are embedded within configurations of non-target or distractor items. If the position of the target relative to the distractors is invariant, such spatial contingencies are implicitly learned and come to guide visual scanning ("contextual cueing"). However, the effectiveness of contextual cueing depends heavily on the consistency between bottom-up perceptual input and context memory:following configural learning, re-locating targets to an unexpected location within an unchanged distractor context completely abolishes contextual cueing, and gains deriving from the invariant context recover only very slowly with increasing exposure to the changed displays. The current study induces variations of the local target context, i.e., item density, to investigate the relation between this factor and contextual adaptation. The results showed that learned contextual cues can be adapted quickly if the target is re-positioned to a sparse local distractor context (consisting of 1 neighbouring non-target item), as compared to no adaptation with a dense context (with 3 surrounding non-targets). This suggests that contextual adaptation is modulated by spatial factors and is not per se limited by order effects in the learning process.

Influences of luminance contrast and ambient lighting on visual context learning and retrieval

“…Interestingly, repeated contexts could still be learned in local search mode, as evidenced by the significant cueing effect manifesting already in the first block of the transfer session, after the switch of the display contrast switched from low to high. This is consistent with previous findings (Annac, Conci, Müller, & Geyer, 2017 ; Zang et al, 2015 ) emphasizing the role of local target–distractor associations in contextual learning.…”

Distributed attention beats the down-side of statistical context learning in visual search

“…Thus this indicates that the efficient adaptation in the distributed attention mode was not simply due to some observers exhibiting "late learning" (Zellin et al, 2013a). Although this pattern essentially replicates previous studies, which also failed to find effective contextual adaptation (e.g., Annac et al, 2017;Zellin et al, 2014), the examination of observers' oculomotor behavior yielded some additional, new insights into the processes of context learning and adaptation: successful contextual learning was characterized by fewer fixations and shorter saccades, while adaptation to the relocated target again resulted in longer saccades. Furthermore, the magnitude of the RT cueing effect was significantly correlated with fixation number (and saccade amplitude).…”

“…Specifically, distributed attention appears to facilitate processing of the whole spatial array, thus providing a context representation that is rather flexible and which can be updated readily. By contrast, when attention is focused, which is the typical, “default” processing mode in search tasks that require close scrutiny of individual items (as in the present letter search task), then changes of the target location are not easily integrated in the existing context memory representation, as evidenced by a lack of contextual adaptation (see also Annac et al, 2017 ; Makovski & Jiang, 2010 ; Manginelli & Pollmann, 2009 ; Zellin et al, 2013a ). However, this does not mean that there is no context adaptation with focused search at all.…”

“…Whereas contextual cueing demonstrates that the learning of statistical regularities provides a powerful means for facilitating search guidance in familiar contexts, the updating of previously acquired context memories has been shown to be far less efficient and rather inflexible ( Manginelli & Pollmann, 2009 ; Makovski & Jiang, 2010 ; Conci, Sun, & Müller, 2011 ; Zellin, Conci, von Mühlenen, & Müller, 2013a ; Annac, Conci, Müller, Geyer, 2017 ). For instance, Zellin, von Mühlenen, Müller, and Conci (2014) had their observers perform a search task that was divided into a learning and test phase.…”

Predictive visual search: Role of environmental regularities in the learning of context cues

“…To equate target location repetition effects between the two types of displays, the target appeared equally often at each of 16 possible locations throughout the experiment: eight locations were used for repeated displays and the remaining eight for nonrepeated displays. Furthermore, item density was controlled for each display, presenting three search items per quadrant (see Annac, Conci, Müller, & Geyer, 2017). The orientation of the target in a given repeated display was selected at random on each trial, whereas the nontargets were held constant across repetitions (cf.…”