Abstract
Objects that are semantically related to the visual scene context are typically better recognized than unrelated objects. While context effects on object recognition are well studied, the question which particular visual information of an object’s surroundings modulates its semantic processing is still unresolved. Typically, one would expect contextual influences to arise from high-level, semantic components of a scene but what if even low-level features could modulate object processing? Here, we generated seemingly meaningless textures of real-world scenes, which preserved similar summary statistics but discarded spatial layout information. In Experiment 1, participants categorized such textures better than colour controls that lacked higher-order scene statistics while original scenes resulted in the highest performance. In Experiment 2, participants recognized briefly presented consistent objects on scenes significantly better than inconsistent objects, whereas on textures, consistent objects were recognized only slightly more accurately. In Experiment 3, we recorded event-related potentials and observed a pronounced mid-central negativity in the N300/N400 time windows for inconsistent relative to consistent objects on scenes. Critically, inconsistent objects on textures also triggered N300/N400 effects with a comparable time course, though less pronounced. Our results suggest that a scene’s low-level features contribute to the effective processing of objects in complex real-world environments.
Highlights
Objects typically do not appear randomly in their surroundings
What information of a scene is sufficient to modulate the semantic processing of objects in their visual context and drives the consistency effect? It is conceivable that scene context influences object processing in two different ways: One way to boost object recognition would be to use existing knowledge of frequent object co-occurrences
Binary single trial responses per participant were submitted to a generalized linear mixed-effects model (GLMM)
Summary
Objects typically do not appear randomly in their surroundings. The regularity of object-scene co-occurrences helps us to give meaning quickly to our visual world, for instance, by predicting that the item on the kitchen counter is a mug and not a roll of toilet paper. Even very briefly presented naturalistic scenes modulate object processing such that objects placed on task-irrelevant background images are named more accurately if they are related compared to unrelated with the background scenes2 — which we refer to as semantic consistency effect. The visual system accurately estimates the average size and orientation of a set of objects without attending to each of them individually[17] Apart from this item-based information, which requires prior segmentation of the display, one can extract a “summary” of global information from a whole scene, without segmentation. Phase statistics are computed across scales to capture a more thorough three dimensional appearance (for a more detailed description of the parameters, see[21]) These sets of statistics are iteratively imposed on Gaussian noise until the generated texture has similar statistical properties as the input image. The P-S model keeps only the local structure of the input image (e.g., edges and periodic patterns), while its global 3D configuration, if any, is lost[22]
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call