Distractor Semantic Interference in Visual Object Recognition Tasks

In real-life scenarios, objects usually appear not only relative to the background but also in relation to other objects. The semantic relationship between the target and distractor, as well as the consistency between the target and the background, simultaneously influence object search and recognition in scenes. Participants completed a visual search task to locate the target object under different background conditions, with eye-tracking and behavioral data recorded. The results showed that related distractors significantly increased the reaction time, decreased the probability of first fixation on the target area, and required more search scans. Additionally, the presence of a background increased the reaction time and scanning frequency, but scene consistency had no significant effect on search efficiency. These findings indicate that in complex scenes, the semantic relationship between objects has a stronger influence on visual search and recognition than scene consistency. The results underscore the necessity of integrating semantic information into the visual search framework to enhance target recognition efficiency in multi-object scenes. Future research should expand the range of semantic categories and incorporate deep learning algorithms to further explore ways to improve visual search and object recognition in practical applications.