Biology subjects

Bracci, S.

Publications and source records attributed to Bracci, S..

10 recordsLinked to original sources

bioRxiv2026

Object dimensions underlying food representations in visual cortex

Recent work has revealed two food-selective areas in the lateral and ventromedial occipitotemporal cortex (OTC). These studies have shown that food selectivity in these regions cannot be explained by mid-level features like shape, texture, or colour but differences in their representational content remain unclear. Across two fMRI experiments conducted in the same group of participants, we characterized the dimensions underlying food representations in lateral and ventral OTC by examining the contribution of action-related object properties, such as manipulability, relevant to object interaction, and visual features, such as colour and ensemble statistics, relevant to object recognition. Our results reveal a clear dissociation between lateral and ventral OTC, indicating that food representations in these regions reflect distinct computational constraints. In lateral OTC, food representations were primarily associated with action-related properties shared between food and other graspable objects, whereas in ventral OTC, food representations were sensitive to surface object properties, such as colour and ensemble configuration. Consistent with this distinction, lateral OTC showed greater sensitivity to individual objects against distinctive background and responded equally to colour and greyscale stimuli, while ventral OTC exhibited greater sensitivity to coloured stimuli and ensembles with no distinctive background. Finally, topographic artificial neural networks implementing architectural constraints meant to capture OTC spatial organization similarly exhibited two dissociable clusters of food-selective units based on sensitivity to ensemble statistics. Together, these findings suggest that lateral food representations reflect action-relevant properties shared with other manipulable objects, whereas ventral food representations arise from surface-based visual features critical for food identification.

neuroscience↗

bioRxiv2026

Encoding models uncover fine-grained feature selectivity for bodies, hands and tools

Category-selective areas in the occipitotemporal cortex (OTC) are typically characterized by broad tuning, yet neuroimaging suggests a finer-grained organization reflecting distinct computational roles. We combined image-level fMRI with artificial neural network (ANN)-based encoding models to investigate the selectivity and feature sensitivity of category-selective areas in ventral and lateral OTC. Using densely sampled fMRI data in three participants across six sessions, we identified functional dissociations between body, hand, and tool responses at the individual image level. Area-specific encoding models accurately predicted responses to millions of novel images, maintaining clear category preferences. Importantly, comparisons between models trained on areas selective for the same category revealed distinct feature sensitivities consistent with the areas anatomical location and hemispheric lateralization. These findings provide evidence for fine-grained specialization within OTC and demonstrate how ANN-based encoding models can uncover the computational, feature-level basis of category selectivity.

neuroscience↗

bioRxiv2025

Investigating action topography in visual cortex and deep artificial neural networks

High-level visual cortex contains category-selective areas embedded within larger-scale topographic maps like animacy and real-world size. Here, we propose action as a key organizing factor shaping visual cortex topography and assess the ability of topographic deep artificial neural networks (DANNs) in capturing this organization. Using fMRI, we examined responses to images of body-parts and objects with different degrees of action properties. In left lateral occipitotemporal cortex, we identified a topographically-organized action gradient, with overlapping activations for bodies, hands, tools, and manipulable objects along a dorsal-posterior to ventral-anterior axis, culminating at the intersection of body parts and objects exhibiting higher action properties. Multivariate analyses confirmed action as a crucial organizing principle, while shape and animacy dominated ventral occipitotemporal cortex and DANNs, which exhibited no action-based organization. Our proposed action dimension serves as a further organizing principle of object categories, advancing understanding of visual cortex organization and its divergence from DANN-based models.

neuroscience↗

bioRxiv2024

Fine-grained neural coding of bodies and body parts in human visual cortex

The visual image of a human body provides a valuable source of socially relevant information. However, our understanding of the neuronal mechanisms underlying body perception in humans remains limited given the spatiotemporal constraints of functional imaging. Here we recorded multi-unit spiking activity in two neurosurgical patients in or near the extrastriate body area (EBA), a critical region for body perception. Our recordings revealed a strong preference for human bodies over a large range of control stimuli. Notably, this preference was driven by a distinct selectivity for body parts. Moreover, the observed body selectivity generalized to non-photographic depictions of bodies such as silhouettes and stick figures. Overall, our study provides an unprecedented access into the representation of bodies in the human visual cortex to bridge the gap between human neuroimaging and macaque electrophysiology studies, and form a solid basis for computational models of human body processing.

neuroscience↗

bioRxiv2023

The representational dynamics of the animal appearance bias in human visual cortex are indicative of fast feedforward processing

The human visual system has a seemingly unique tendency to interpret zoomorphic objects as animals, not as objects. This animal appearance bias is very strong in the ventral visual pathway as measured through functional magnetic resonance imaging (fMRI), but it is absent in feedforward deep convolutional neural networks. Here we investigate how this bias emerges over time by probing its representational dynamics through multivariate electroencephalography (EEG). The initially activated representations to lookalike zoomorphic objects are very similar to the representations activated by animal pictures and very different from the neural responses to regular objects. Neural responses that reflect the true identity of the zoomorphic objects as inanimate objects are weaker and appear later, as do effects of task context. The strong early emergence of an animal appearance bias strongly supports a feedforward explanation, indicating that lack of recurrence in deep neural networks is not an explanation for their failure to show this bias.

neuroscience↗

bioRxiv2022

A computational understanding of zoomorphic perception in the human brain

It is common to find objects that resemble animals on purpose (e.g., toys). While the perception of such objects as animal-like seems obvious to humans, such "Animal bias" for zoomorphic objects turned out to be a striking discrepancy between the human brain and artificial visual systems known as deep neural networks (DNNs). We provide a computational understanding of the human Animal bias. We successfully induced this bias in DNNs trained explicitly with zoomorphic objects. Alternative training schedules, focusing on previously identified differences between the brain and DNNs, failed to cause an Animal bias. Specifically, we considered the superordinate distinction between animate and inanimate classes, the sensitivity for faces and bodies, the bias for shape over texture, and the role of ecologically valid categories. These findings provide computational support that the Animal bias for zoomorphic objects is a unique property of human perception yet can be explained by human learning history.

neuroscience↗

bioRxiv2021

Object-scene conceptual regularities reveal fundamental differences between biological and artificial object vision

Human vision is still largely unexplained. Computer vision made impressive progress on this front, but it is still unclear to which extent artificial neural networks approximate human object vision at the behavioral and neural levels. Here, we investigated whether machine object vision mimics the representational hierarchy of human object vision with an experimental design that allows testing within-domain representations for animals and scenes, as well as across-domain representations reflecting their real-world contextual regularities such as animal-scene pairs that often co-occur in the visual environment. We found that DCNNs trained in object recognition acquire representations, in their late processing stage, that closely capture human conceptual judgements about the co-occurrence of animals and their typical scenes. Likewise, the DCNNs representational hierarchy shows surprising similarities with the representational transformations emerging in domain-specific ventrotemporal areas up to domain-general frontoparietal areas. Despite these remarkable similarities, the underlying information processing differs. The ability of neural networks to learn a human-like high-level conceptual representation of object-scene co-occurrence depends upon the amount of object-scene co-occurrence present in the image set thus highlighting the fundamental role of training history. Further, although mid/high-level DCNN layers represent the category division for animals and scenes as observed in VTC, its information content shows reduced domain-specific representational richness. To conclude, by testing within- and between-domain selectivity while manipulating contextual regularities we reveal unknown similarities and differences in the information processing strategies employed by human and artificial visual systems. Author SummaryComputational object vision represents the new frontier of brain models, but do current artificial visual systems known as deep convolutional neural networks (DCNNs) represent the world as humans do? Our results reveal that DCNNs are able to capture important representational aspects of human vision both at the behavioral and neural levels. At the behavioral level, DCNNs are able to pick up contextual regularities of objects and scenes thus mimicking human high-level semantic knowledge such as learning that a polar bear "lives" in ice landscapes. At the neural representational level, DCNNs capture the representational hierarchy observed in the visual cortex all the way up to frontoparietal areas. Despite these remarkable correspondences, the information processing strategies implemented differ. In order to aim for future DCNNs to perceive the world as humans do, we suggest the need to consider aspects of training and tasks that more closely match the wide computational role of human object vision over and above object recognition.

neuroscience↗

bioRxiv2020

Neural correlates of hand-object congruency effects during action planning

Selecting hand actions to manipulate an object is affected both by perceptual factors and by action goals. Affordances are associated with the automatic potentiation of motor representations to an object, independent of the goal of the actor. In previous studies, we have demonstrated an influence of the congruency between hand and object orientations on response times when reaching to turn an object, such as a cup. In this study, we investigated how the representation of hand postures triggered by planning to turn a cup were influenced by this congruency effect, in an fMRI scanning environment. Healthy participants were asked to reach and turn a real cup that was placed in front of them either in an upright orientation or upside down. They were instructed to use a hand orientation that was either congruent or incongruent with the cup orientation. As expected, the motor responses were faster when the hand and cup orientations were congruent. There was increased activity in a network of brain regions involving object-directed actions during action planning, which included bilateral primary and extrastriate visual, medial and superior temporal areas, as well as superior parietal, primary motor and premotor areas in the left hemisphere. Specific activation of the dorsal premotor cortex (PMd) was associated with hand-object orientation congruency during planning, and prior to any action taking place. Activity in that area and its connectivity with the lateral occipito-temporal cortex (LOTC) increased when planning incongruent actions. The increased activity in premotor areas in trials where the orientation of the hand was incongruent to that of the object suggests a role in eliciting competing representations specified by hand postures in LOTC.

neuroscience↗

bioRxiv2019

Orthogonal Representations of Object Shape and Category in Deep Convolutional Neural Networks and Human Visual Cortex

Deep Convolutional Neural Networks (CNNs) are gaining traction as the benchmark model of visual object recognition, with performance now surpassing humans. While CNNs can accurately assign one image to potentially thousands of categories, network performance could be the result of layers that are tuned to represent the visual shape of objects, rather than object category, since both are often confounded in natural images. Using two stimulus sets that explicitly dissociate shape from category, we correlate these two types of information with each layer of multiple CNNs. We also compare CNN output with fMRI activation along the human visual ventral stream by correlating artificial with biological representations. We find that CNNs encode category information independently from shape, peaking at the final fully connected layer in all tested CNN architectures. Comparing CNNs with fMRI brain data, early visual cortex (V1) and early layers of CNNs encode shape information. Anterior ventral temporal cortex encodes category information, which correlates best with the final layer of CNNs. The interaction between shape and category that is found along the human visual ventral pathway is echoed in multiple deep networks. Our results suggest CNNs represent category information independently from shape, much like the human visual system.

neuroscience↗

bioRxiv2017

The ventral visual pathway represents animal appearance over animacy, unlike human behavior and deep neural networks

Recent studies showed agreement between how the human brain and neural networks represent objects, suggesting that we might start to understand the underlying computations. However, we know that the human brain is prone to biases at many perceptual and cognitive levels, often shaped by learning history and evolutionary constraints. Here we explore one such bias, namely the bias to perceive animacy, and used the performance of neural networks as a benchmark. We performed an fMRI study that dissociated object appearance (how an object looks like) from object category (animate or inanimate) by constructing a stimulus set that includes animate objects (e.g., a cow), typical inanimate objects (e.g., a mug), and, crucially, inanimate objects that look like the animate objects (e.g., a cow-mug). Behavioral judgments and deep neural networks categorized images mainly by animacy, setting all objects (lookalike and inanimate) apart from the animate ones. In contrast, activity patterns in ventral occipitotemporal cortex (VTC) were strongly biased towards object appearance: animals and lookalikes were similarly represented and separated from the inanimate objects. Furthermore, this bias interfered with proper object identification, such as failing to signal that a cow-mug is a mug. The bias in VTC to represent a lookalike as animate was even present when participants performed a task requiring them to report the lookalikes as inanimate. In conclusion, VTC representations, in contrast to neural networks, fail to veridically represent objects when visual appearance is dissociated from animacy, probably due to a biased processing of visual features typical of animate objects.

neuroscience↗