Face Recognition and Perception

This study compares face and body processing in humans and deep neural networks, finding both segregated and integrated representations across visual cortex regions.

The study examines how infants' neural responses to faces change with age and experience, showing differences in brain activity for familiar and unfamiliar social groups.

The paper shows that prior scene context influences rapid face detection by shifting reliance from sensory input to spatial expectations.

This study shows that self-supervised learning in neural networks can replicate human category-selective brain regions' representational signatures.

The paper shows humans can follow eye-gaze cues from both human and rhesus monkey avatars equally well, challenging the idea that ocular morphology limits gaze-following ability.

The paper investigates how face-likeness affects memory and neural representations, finding that non-face-like houses are better remembered and that representational reliability in the FFA correlates with memory performance.

The paper investigates how different types of information (visual, identity, semantic) drive ERP components (N170, N250, SFE) in familiar face recognition using a sequential familiarization paradigm.

The paper identifies a shared neural representation for gender across faces, bodies, and objects in the human brain, highlighting a region in the right middle temporal gyrus that encodes gender independently of stimulus category.

The paper shows how task demands dynamically structure feature selection, routing, and integration in the human brain using dynamic faces and MEG data.

This paper introduces EmoFB, a deep neural network model that simulates emotional perception by integrating top-down feedback signals, showing improved recognition and alignment with human brain responses.

The paper investigates how low-dimensional frontal feedback resolves visual ambiguity in the human brain using fMRI, EEG, and computational models.

This study uses hyperalignment to estimate individualized functional topographies in developmental prosopagnosia without localizer scans, showing high fidelity and cross-group generalization.

This study investigates how threat-relevance modulates early visual processing and autonomic responses to faces, finding distinct roles for magnocellular pathways and conscious anticipation.

This study shows that exposing deep neural networks to diverse facial groups reduces bias and improves cross-group face recognition, aligning with human behavior.

The paper shows that reading direction influences left-side perceptual bias and gaze patterns in face matching tasks.

The paper introduces a reinforcement learning model for face perception that incorporates environmental feedback, showing it matches supervised and unsupervised models in capturing neural responses to faces.

This study shows that face-deprived monkeys develop face selectivity later but retain lasting effects of early deprivation on neural representations.

This study trains two ANNs on different visual diets to show how specialized vs. broad experience shapes face-processing capabilities and generalization.

The paper shows prior scene context influences neural dynamics during rapid face detection, with early facilitation and later compensatory processing when context is absent.

The paper shows that activation in category-selective regions of the occipito-temporal cortex predicts categorisation behaviour in face, body, and scene classification tasks.

This study investigates how VTC features influence object manifold separability using ANN modeling, fMRI, and MEG, showing dynamic geometric adjustments in neural representations.

The paper investigates whether serial dependence arises from physical or categorical representations using behavioral and EEG data on facial expressions.

This paper identifies a right-lateralized neural core and adaptive left-hemisphere components involved in non-face visual expertise across domains using meta-analysis of fMRI studies.

The paper investigates how emotional expressions are represented in brain networks for facial and vocal signals, finding shared neural codes and multisensory integration across regions.

This study compares how convolutional neural networks and Vision Transformers align with human eye movement patterns during facial expression recognition, finding that ViT better captures sequential fixation strategies.

The paper investigates whether ensemble encoding of facial expressions differs between dynamic and static faces, finding that dynamic faces show reduced accuracy with larger ensembles.

This paper introduces a developmentally inspired model for face recognition that learns continuously through generative memory replay, achieving performance comparable to batch-trained DCNNs.

This study identifies distinct functional roles of human amygdala subregions in fearful-expression recognition through intracranial EEG and stimulation, revealing bidirectional interactions and a temporal hierarchy.

This study uses VASO 7T layer fMRI and MVPA to show that face information is structured in the middle layer of the fusiform face area during face processing.

This study investigates how central vision loss alters internal emotion representations using a genetic algorithm paradigm and three-dimensional face models, finding significant group differences in emotion expression.