Biology subjects

Peelen, M. V.

Publications and source records attributed to Peelen, M. V..

At least 19 recordsLinked to original sources

bioRxiv2026

Turning an object into a scene: buildings activate scene-selective visual cortex independently of visual features

Human visual cortex contains regions that selectively respond to both scenes and large objects, particularly buildings. The cortical overlap between buildings and scenes has been attributed to shared visual features (e.g., cardinal orientations, rectilinearity). Alternative accounts propose that buildings may also activate scene representations indirectly, independently of specific visual features, for example because buildings evoke a sense of space. Here, we tested for such feature-independent activation by comparing EEG and fMRI responses in human participants (both sexes) to buildings and visually-matched boxes, and relating these responses to scene-selective responses. Buildings and boxes were matched, across exemplars, using image-based metrics, deep neural networks, and a perceptual similarity task. Time-resolved EEG decoding showed that buildings and boxes evoked discriminable responses from 360ms post-stimulus onset, incompatible with feedforward visual feature processing. Importantly, the building-box classifier generalized to discriminate scenes from chairs, providing EEG evidence for a representational overlap between buildings and scenes. Temporal generalization analyses further showed that the late building-selective response corresponded to an earlier scene-selective response, with a temporal offset of [~]130ms. Finally, ultra-fast fMRI (TR=140ms) revealed that these findings were mirrored in the response of the scene-selective parahippocampal place area (PPA), which similarly showed a feature-independent building-selective response that was delayed and prolonged relative to the scene-selective response. These results clarify the nature of building selectivity in visual cortex by showing that this selectivity can arise independently of visual features, putatively reflecting associative processes between buildings and scenes (or space). Significance StatementThe human ventral stream contains regions responding selectively to behaviorally-relevant categories such as places, faces, and words. Contrasting views attribute this selectivity either to the feedforward processing of visual features or to top-down input from domain-specific networks. Here, we show that neural selectivity for buildings persists in the absence of category-diagnostic visual features. EEG decoding showed that this high-level building response emerged late during object processing and corresponded to an earlier scene-selective response. Ultra-fast fMRI revealed that these findings were mirrored in the response of the scene-selective parahippocampal place area (PPA). These results show that selectivity for some categories can be decoupled from visual feature processing, reflecting associative processes that follow object recognition.

neuroscience↗

bioRxiv2026

Behavioral and causal evidence for object-based scene recognition in visual cortex

Our visual environment can be parsed into objects and scenes, a distinction that is reflected in the organization of the human visual cortex. Previous research has shown that object and scene perception nevertheless closely interact, such that scenes influence object perception and objects influence scene perception. It remains unclear, however, whether and how objects that are not inherently diagnostic of their surroundings aid the recognition of poorly visible scenes (e.g., a person standing in a dark living room). Here, in three behavioral experiments, we show that participants made more accurate indoor/outdoor judgments when degraded scene photographs were presented together with an object than when the scene or the object was shown alone, even though the same object categories appeared in indoor and outdoor scenes. This object-driven benefit vanished once scene structure was removed through phase scrambling and was reduced when objects appeared in physically inconsistent locations within the scenes. These results suggest that objects in consistent locations (e.g., a person standing on a floor) disambiguate scene layout. Finally, in a pre-registered transcranial magnetic stimulation (TMS) study (N = 48), we provide causal evidence that the object-selective lateral occipital cortex (LOC) supports scene categorization when scene layout is disambiguated by within-scene objects. Stimulation of the LOC, particularly at 260-300 ms after stimulus onset, selectively disrupted object-based scene recognition. Together, these findings demonstrate that objects facilitate the read-out of the surrounding space in service of efficient scene recognition. Significance StatementUnderstanding how scene and object processing interact for efficient recognition is a key question in natural vision. Research has long emphasized how surrounding scenes help us identify objects, yet the reverse - how objects shape the perception of scenes - has received little attention. In the dark, does a glimpse of a floating boat tell us we are looking at a lake? In this study we demonstrate that a single object helps people recognize hardly visible scenes. This benefit required intact scene structure and depended on where the object appeared in the scene. In addition, object selective visual cortex was causally related to this benefit. Together, these findings show that objects visual appearance can be used to better understand our surroundings.

neuroscience↗

bioRxiv2025

Contextual associations impact visual search across multiple processing stages

Growing evidence indicates that visual search is influenced not only by perceptual but also by semantic information. Among such semantic influences, categorical effects have been extensively studied. By contrast, the influence of contextual associations, that is, how objects commonly co-occur within specific scene contexts, remains less understood. Here, we orthogonally manipulated contextual and categorical relationships between target and distractor objects in a visual search task, while controlling for perceptual target-distractor similarity. By applying linear regression analyses, we modelled behavioral responses and event-related potentials (ERPs) from contextual, categorical, and perceptual target-distractor similarity. We found that contextual associations guide attention by interacting with categorical information (modulating the N2pc) and ultimately shape behavioral search performance jointly with perceptual and categorical similarity. These findings demonstrate that contextual object associations can guide attention, thereby supporting efficient search in complex visual environments.

neuroscience↗

bioRxiv2025

Narrative Context Shifts Gaze from Visual to Semantic Salience

Humans make over a hundred thousand eye movements daily to gather visual information. But what determines where we look? Current computational models typically link gaze behaviour to visual features of isolated images, but we know that eye movements are also strongly shaped by cognitive goals: Observers gather information that helps them to understand, rather than just represent, the world. Within this framework, observers should focus more on information that updates ones understanding of the environment, and less on what is purely visually salient. Here we tested this hypothesis using a free-viewing paradigm of narratives where we experimentally manipulated the meaningfulness of temporal context by either presenting pictures in a coherent, i.e. correct, order, or in a temporally shuffled order. We developed a novel approach to quantify which visual information is semantically salient (i.e., important for understanding): we separately obtained language narratives for images in stories, and computed the contextual surprisal of visual objects using a large language model. The ability of this semantic saliency model in explaining gaze behaviour was compared to a state-of-the art model of visual saliency (DeepGaze-II). We found that individuals looked relatively more often and more quickly at semantically salient objects when images were presented in coherent compared to shuffled order. In contrast, visual salience did not better account for gaze behaviour in coherent than shuffled order. These findings highlight how internal contextual models guide visual sampling and demonstrate that language models could offer a powerful tool for capturing gaze behavior in richer, meaningful settings.

neuroscience↗

bioRxiv2025

Spatiotemporal representations of contextual associations for real-world objects

In the real world, objects always appear in context. Many objects are reliably associated with a certain scene context (e.g., pots appear in kitchens) and other objects that appear in the same context (e.g., pans appear together with pots). Previous neuroimaging work suggests that such contextual associations shape the neural representation of isolated objects even in the absence of the scene context. Yet, three key questions remain unanswered: (1) How do representations of contextual associations relate to perceptual and categorical representation in visual cortex, (2) how do they emerge across time, and (3) how are they mechanistically implemented? To answer these questions, we recorded fMRI and EEG while participants (human, both sexes) viewed isolated objects stemming from two scene contexts. Multivariate pattern analysis on the neural data revealed that objects from the same context were coded more similarly than objects from different contexts in object-selective LOC and scene-selective PPA, even when systematically controlling for perceptual and categorical similarities. Such contextual relation representations emerged relatively late during visual processing (i.e., after perceptual and categorical representations), specifically in the anterior PPA, and likely through a mixture of object-to-object and object-to-scene associations. Together, our results demonstrate that contextual relation representations emerged for isolated objects, and without a task that encourages their formation, suggesting that objects automatically activate context frames that support visual cognition in real-world environments.

neuroscience↗

bioRxiv2025

Unpacking similarity effects in visual memory search: categorical, semantic, and visual contributions

Visual memory search involves comparing a probe item against multiple memorized items. Previous work has shown that distractor probes from a different object category than the objects in the memory set are rejected more quickly than distractor probes from the same category. Because objects belonging to the same superordinate category usually share both visual and semantic features compared with objects of different categories, it is unclear whether the category effects reported in previous studies reflected visual and/or semantic target-distractor similarity. Here, we employed old/new recognition tasks to examine the role of categorical, semantic, and visual similarity in short- and long-term memory search. Participants (N=64) performed visual long-term memory (LTM) or short-term memory (STM) search tasks involving animate and inanimate objects. Trial-wise RT variability to distractor probes in LTM and STM search was modelled using regression analyses that included predictors capturing categorical target-distractor similarity (same or different category), semantic target-distractor similarity (from a distributional semantic model), and visual target-distractor similarity (from a deep neural network). We found that categorical, semantic, and visual similarity all explained unique variance in trial-wise memory search performance. However, their respective contributions varied with memory set size and task, with STM performance being relatively more strongly influenced by visual and categorical similarity and LTM performance being relatively more strongly influenced by semantic similarity. These results clarify the nature of the representations used in memory search and reveal both similarities and differences between search in STM and LTM.

neuroscience↗

bioRxiv2025

Object representations drive emotion schemas across a large and diverse set of daily-life scenes

The rapid emotional evaluation of objects and events is essential in daily life. While visual scenes reliably evoke emotions, it remains unclear whether emotion schemas evoked by daily-life scenes depend on object processing systems or are extracted independently. To explore this, we collected emotion ratings for 4913 daily-life scenes from 300 participants, and predicted these ratings from representations in deep neural networks and fMRI activity patterns in visual cortex. AlexNet, an object-based model, outperformed EmoNet, an emotion-based model, in predicting emotion ratings for everyday scenes, while EmoNet excelled for explicitly evocative stimuli. Emotion information was processed hierarchically within the object recognition system, consistent with the visual cortexs organization. Activity patterns in the lateral occipital complex (LOC), an object-selective region, reliably predicted emotion ratings and outperformed other visual regions. These findings suggest that emotion processing in everyday scenes follows visual object recognition, with additional mechanisms engaged when object content is uninformative.

neuroscience↗

bioRxiv2025

Knowledge of effort modulates visual memory biases for body postures

The visual memory of others postures has been proposed to be shaped by knowledge and expectations. For example, the visual memory of a lifted arm was recently shown to be biased downward, suggesting that observers predicted the upcoming state of the arm based on knowledge of the effort required to hold the arm up against gravity. Alternatively, the downward bias for body postures could reflect an automatic normalization towards the most frequently observed arm position, with arms more often observed in a low position. Here, in three experiments, we provide evidence that the downward bias is flexibly modulated by knowledge of effort. In Experiment 1, we found a stronger downward bias for arm postures that are relatively effortful (lifting an arm above the shoulders while standing) compared to arm postures that are less effortful (lifting an arm above the chest while lying down). In Experiment 2, we found a stronger downward bias when the actor was standing (viewed from the side) than when the actor was lying down (viewed from above), even though the arm postures were visually identical. Moreover, dividing attention during the encoding stage reduced the bias, showing that attentive processing of the stimulus was required for the bias to emerge. Finally, in Experiment 3, we found that executing the observed posture during the visual memory task did not further increase the downward bias. Together, these findings demonstrate a high-level cognitive influence on the visual memory for body postures.

neuroscience↗

bioRxiv2024

A body detection inversion effect revealed by a large-scale inattentional blindness experiment

As a social species, humans preferentially attend to the faces and bodies of other people. Previous research revealed specialized cognitive mechanisms for processing human faces and bodies. For example, upright person silhouettes are more readily found than inverted silhouettes in visual search tasks. It is unclear, however, whether these findings reflect a top-down attentional bias to social stimuli or bottom-up sensitivity to visual cues signaling the presence of other people. Here, we tested whether the upright human form is preferentially detected in the absence of attention. To rule out influences of top-down attention and expectation, we conducted a large-scale single-trial inattentional blindness experiment on a diverse sample of naive participants (N=13.539). While participants were engaged in judging the length of a cross at fixation, we briefly presented an unexpected silhouette of a person or a plant next to the cross. Subsequently, we asked whether participants noticed anything other than the cross. Results showed that silhouettes of people were more often noticed than silhouettes of plants. Crucially, upright person silhouettes were also more often detected than inverted person silhouettes, despite these stimuli being identical in their low-level visual features. These results were replicated in a second experiment involving headless person silhouettes. Finally, capitalizing on the exceptionally large and diverse sample, further analyses revealed strong detection differences across age and gender. These results indicate that the visual system is tuned to the form of the upright human body, allowing for the quick detection of other people even in the absence of attention.

neuroscience↗

bioRxiv2024

The neural time course of size constancy in natural scenes

Accurate real-world size perception relies on size constancy, a mechanism that integrates an objects retinal size with distance information. The neural time course of extracting pictorial distance cues from scenes and integrating them with retinal size information - a process referred to as scene-based size constancy - remains unknown. In two experiments, participants viewed objects with either large or small retinal sizes, presented at near or far distances in outdoor scene photographs, while performing an unrelated one-back task. We applied multivariate pattern analysis (MVPA) to time-resolved EEG data to decode the retinal size of large versus small objects, depending on their distance (near versus far) in the scenes. The objects were either perceptually similar in size (large-near versus small-far) or perceptually dissimilar in size (large-far versus small-near), reflecting size constancy. We found that the retinal size of objects could be decoded from 80 ms after scene onset onwards. Distance information modulated size decoding at least 120 ms later: from 200 ms after scene onset when objects were fixated, and from 280 ms when objects were viewed in the periphery. These findings reveal the neural time course of size constancy based on pictorial distance cues in natural scenes.

neuroscience↗

bioRxiv2024

Scene context and attention independently facilitate MEG decoding of object category

Many of the objects we encounter in our everyday environments would be hard to recognize without any expectations about these objects. For example, a distant silhouette may be perceived as a car because we expect objects of that size, positioned on a road, to be cars. Reflecting the influence of such expectations on visual processing, neuroimaging studies have shown that when objects are poorly visible, expectations derived from scene context facilitate the representations of these objects in visual cortex from around 300 ms after scene onset. The current magnetoencephalography (MEG) study tested whether this facilitation occurs independently of attention and task relevance. Participants viewed degraded objects alone or within their original scene context while they either attended the scenes (attended condition) or the fixation cross (unattended condition), temporally directing attention away from the scenes. Results showed that at 300 ms after stimulus onset, multivariate classifiers trained to distinguish clearly visible animate vs inanimate objects generalized to distinguish degraded objects in scenes better than degraded objects alone, despite the added clutter of the scene background. Attention also modulated object representations at this latency, with better category decoding in the attended than the unattended condition. The modulatory effects of context and attention were independent of each other. Finally, data from the current study and a previous study were combined (N=51) to provide a more detailed temporal characterization of contextual facilitation. These results extend previous work by showing that facilitatory scene-object interactions are independent of the specific task performed on the visual input.

neuroscience↗

bioRxiv2024

Electrophysiological correlates of visual memory search

In everyday life, we frequently engage in hybrid search, where we look for multiple items stored in memory (e.g., a mental shopping list) in our visual environment. Across three experiments, we used event-related potentials to better understand the contributions of visual working memory (VWM) and long-term memory (LTM) during the memory search component of hybrid search. Experiments 1 and 2 demonstrated that the FN400 - an index of LTM recognition - and the CDA - an index of VWM - increased with memory set size (target load), suggesting that both VWM and LTM are involved in memory search, even when memory load exceeds capacity limitations of VWM. In Experiment 3, we used these electrophysiological indices to test how categorical similarity of targets and distractors affects memory search. The CDA and FN400 were modulated by memory set size only if items resembled targets. This suggests that dissimilar distractor items can be rejected before eliciting a memory search. Together, our findings demonstrate the interplay of VWM and LTM processes during memory search for multiple targets.

neuroscience↗

bioRxiv2024

Preparatory activity during visual search reflects attention-guiding objects rather than search targets

Efficient behavior requires the rapid attentional selection of task-relevant objects. Preparatory activity of target-selective neurons in visual cortex is thought to support attentional selection, guiding spatial attention and favoring processing of target-matching input. However, naturalistic searches are often guided by non-targets, including target-associated "anchor" objects. For instance, when looking for a pen, we may direct our attention to the office desk on which we expect to find it. Here, using fMRI and eyetracking in a context-guided search task, we tested whether preparatory activity in visual cortex reflected the target, the guiding anchor object, or both. Participants learned associations between targets and anchors, reversing across two scene contexts, before searching for these targets. Participants first fixations were reliably guided by the associated anchor. Preparatory activity in lateral occipital cortex (LOC), and right intraparietal sulcus (IPS), represented the target-associated anchor rather than the target. These results shed light on the neural basis of context-guided search in structured environments.

neuroscience↗

bioRxiv2023

Category-based attention facilitates memory search

We often need to decide whether the object we look at is also the object we look for. When we look for one specific object, this process can be facilitated by preparatory feature-based attention. However, when we look for multiple objects at the same time (e.g., the products on our shopping list) such a strategy may no longer be possible, as research has shown that we can actively prepare to detect only one object at a time. Therefore, looking for multiple objects may additionally involve search in long-term memory, slowing down decision making. Interestingly, however, previous research has shown that memory search can be very efficient when distractor objects are from a different category than the items in the memory set. Here, using EEG, we show that this efficiency is supported by top-down attention at the category level. In Experiment 1, human participants (both sexes) performed a memory search task on individually presented objects of the same or different category as the objects in the memory set. We observed category-level attentional modulation of distractor processing from [~]150 ms after stimulus onset, expressed both as an evoked response modulation and as an increase in decoding accuracy of same-category distractors. In Experiment 2, memory search was performed on two concurrently presented objects. When both objects were distractors, spatial attention (indexed by the N2pc component) was directed to the object that was of the same category as the objects in the memory set. Together, these results demonstrate how attention can facilitate memory search. Significance statementWhen we are in the supermarket, we repeatedly decide whether a product we look at (e.g., a banana) is on our memorized shopping list (e.g., apples, oranges, kiwis). This requires searching our memory, which takes time. However, when the product is of an entirely different category (e.g., dairy instead of fruit), the decision can be made quickly. Here, we used EEG to show that this between-category advantage in memory search tasks is supported by top-down attentional modulation of visual processing: The visual response evoked by distractor objects was modulated by category membership, and spatial attention was quickly directed to the location of within-category (vs. between-category) distractors. These results demonstrate a close link between attention and memory.

neuroscience↗

bioRxiv2023

Predicting cued and oddball visual search performance from neural representational similarity

Capacity limitations in visual tasks can be observed when the number of task-related objects increases. An influential idea is that such capacity limitations are determined by competition at the neural level: two objects that are encoded by shared neural populations interfere more in behavior (e.g., visual search) than two objects encoded by separate neural populations. However, the neural representational similarity of objects varies across brain regions and across time, raising the question of where and when competition determines task performance. Furthermore, it is unclear whether the association between neural representational similarity and task performance is common or unique across tasks. Here, we used neural representational similarity derived from fMRI, MEG, and deep neural networks (DNN) to predict performance on two visual search tasks involving the same objects and requiring the same responses but differing in instructions: cued visual search and oddball visual search. Separate groups of human participants (both sexes) viewed the individual objects in neuroimaging experiments to establish the neural representational similarity between those objects. Results showed that performance on both search tasks could be predicted by neural representational similarity throughout the visual system (fMRI), from 80 msec after onset (MEG), and in all DNN layers. Stepwise regression analysis, however, revealed task-specific associations, with unique variability in oddball search performance predicted by early/posterior neural similarity, and unique variability in cued search task performance predicted by late/anterior neural similarity. These results reveal that capacity limitations in superficially similar visual search tasks may reflect competition at different stages of visual processing. Significance StatementVisual search for target objects is slowed down by the presence of distractors, but not all distractors are equally distracting - the more similar a distractor is to the target, the more it slows down search. Here, we used fMRI, MEG, and a deep neural network to reveal where, when, and how neural similarity between targets and distractors predicts visual search performance across two search tasks: oddball visual search (locating the different-looking object) and cued visual search (locating the cued object). Results also revealed brain regions, time points, and feature levels that predicted task-unique performance. These results provide a neural basis for similarity theories of visual search and show that this neural basis differs across visual search tasks.

neuroscience↗

bioRxiv2023

Better together: Objects in familiar constellations evoke high-level representations of real-world size

High-level vision is frequently studied at the level of either individual objects or whole scenes. An intermediate level of visual organisation that has received less attention is the "object constellation" - a familiar configuration of contextually-associated objects (e.g., plate + spoon). Recent behavioural studies have shown that information from multiple objects can be integrated to support observers high-level understanding of a "scene" and its constituent objects. Here we used EEG in human participants (both sexes) to test when the visual system integrates information across objects to support recognition, using representations of objects real-world size as a proxy for recognition. We briefly presented masked object constellations consisting of object silhouettes of either large (e.g., chair + table) or small (e.g., plate + spoon) real-world size, while independently varying retinal size. As a control, observers also viewed each silhouette in isolation. If object context facilitates object recognition, real-world size should be inferred more effectively when the objects appear in their contextually-associated pairs than in isolation, leading to the emergence of real-world size information in multivariate EEG patterns. Representational similarity analysis revealed that neural activity patterns captured information about the real-world size of object constellations from [~]200 ms after stimulus onset. This representation was stronger for, and specific to, object pairs as compared to single objects, and remained significant after regressing out visual similarity models derived from computational models. These results provide evidence for inter-object facilitation of visual processing, leading to a qualitatively different high-level representation of object pairs than single objects. Significance StatementThis study used electroencephalography decoding to reveal the neural timecourse of inter-object facilitation present for contextually-associated groups of objects (e.g., chair + table). Although ubiquitous in daily life, the object constellation level of representation has rarely been examined compared to isolated objects or entire scenes. By shedding new light on facilitatory interactions between objects, arising before 200ms of visual processing, our results provide insight into the continuum along which objects and scenes exist. At the same time, this work advances the current understanding of the neural basis of real-world size, using strict visual controls to show that inferred information about objects spatial scale in the real world emerges around 200 ms after stimulus onset.

neuroscience↗

bioRxiv2023

Objects sharpen visual scene representations: evidence from MEG decoding

Real-world scenes consist of objects, defined by local information, and scene background, defined by global information. While objects and scenes are processed in separate pathways in visual cortex, their processing interacts. Specifically, previous studies have shown that scene context makes blurry objects look sharper, an effect that can be observed as a sharpening of object representations in visual cortex from around 300 ms after stimulus onset. Here, we use MEG to show that objects can also sharpen scene representations, with the same temporal profile. Photographs of indoor (closed) and outdoor (open) scenes were blurred such that they were difficult to categorize on their own but easily disambiguated by the inclusion of an object. Classifiers were trained to distinguish MEG response patterns to intact indoor and outdoor scenes, presented in an independent run, and tested on degraded scenes in the main experiment. Results revealed better decoding of scenes with objects than scenes alone and objects alone from 300 ms after stimulus onset. This effect was strongest over left posterior sensors. These findings show that the influence of objects on scene representations occurs at similar latencies as the influence of scenes on object representations, in line with a common predictive processing mechanism.

neuroscience↗

bioRxiv2022

Prior Knowledge Biases the Perception of Body Postures

Body postures provide information about others actions, intentions, and emotional states. Little is known about how postures are represented in the brains visual system. Considering our extensive visual and motor experience with body postures, we hypothesized that priors derived from this experience may systematically bias visual body posture representations. We examined two priors: gravity and biomechanical constraints. Gravity pushes lifted body parts downwards, while biomechanical constraints limit the range of possible postures (e.g., an arm raised far behind the head cannot go down further). Across three experiments (N=246) we probed participants visual memory of briefly presented postures using change discrimination and adjustment tasks. Results showed that lifted arms were misremembered as lower and as more similar to the nearest biomechanically plausible postures. Inverting the body stimuli eliminated both biases, ruling out visual confounds. These findings show that visual memory representations of body postures are modulated by a combination of category-general and category-specific priors.

neuroscience↗