Semantic and Visual Determinants of Face Recognition in a Prosopagnosic Patient

Prosopagnosia is the neuropathological inability to recognize familiar people by their faces. It can occur in isolation or can coincide with recognition deficits for other nonface objects. Often, patients whose prosopagnosia is accompanied by object recognition difficulties have more trouble identifying certain categories of objects relative to others. In previous research, we demonstrated that objects that shared multiple visual features and were semantically close posed severe recognition difficulties for a patient with temporal lobe damage. We now demonstrate that this patient's face recognition is constrained by these same parameters. The prosopagnosic patient ELM had difficulties pairing faces to names when the faces shared visual features and the names were semantically related (e.g., Tonya Harding, Nancy Kerrigan, and Josée Chouinard— three ice skaters). He made tenfold fewer errors when the exact same faces were associated with semantically unrelated people (e.g., singer Celine Dion, actress Betty Grable, and First Lady Hillary Clinton). We conclude that prosopagnosia and co-occurring category-specific recognition problems both stem from difficulties disambiguating the stored representations of objects that share multiple visual features and refer to semantically close identities or concepts.

Download Full-text

Brain-like functional specialization emerges spontaneously in deep neural networks

10.1101/2021.07.05.451192 ◽

2021 ◽

Author(s):

Katharina Dobs ◽

Julio Martinez ◽

Alexander J.E. Kell ◽

Nancy Kanwisher

Keyword(s):

Neural Networks ◽

Face Recognition ◽

Object Recognition ◽

Cognitive Neuroscience ◽

Visual Features ◽

Functional Specialization ◽

Quarter Century ◽

Cortical Regions ◽

High Degree ◽

The Brain

The last quarter century of cognitive neuroscience has revealed numerous cortical regions in humans with distinct, often highly specialized functions, from recognizing faces to understanding language to thinking about what other people are thinking. But it remains unclear why the cortex exhibits this high degree of functional specialization in the first place. Here, we consider the case of face perception, using artificial neural networks to test the hypothesis that functional segregation of face recognition in the brain reflects the computational requirements of the task. We find that networks trained on generic object recognition perform poorly on face recognition and vice versa, and further that networks optimized for both tasks spontaneously segregate themselves into separate systems for faces and objects. Thus, generic visual features that suffice for object recognition are apparently suboptimal for face recognition and vice versa. We then show functional segregation to varying degrees for other visual categories, revealing a widespread tendency for optimization (without built-in task-specific inductive biases) to lead to functional specialization in machines and, we conjecture, also brains.

Download Full-text

View-tuned and view-invariant face encoding in IT cortex is explained by selected natural image fragments

Scientific Reports ◽

10.1038/s41598-021-86842-7 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Yunjun Nam ◽

Takayuki Sato ◽

Go Uchida ◽

Ekaterina Malakhova ◽

Shimon Ullman ◽

...

Keyword(s):

Face Recognition ◽

Visual Pathway ◽

Visual Features ◽

Natural Image ◽

Face Representation ◽

Low Level ◽

Ventral Visual Pathway ◽

Neural Substrate ◽

The Face

AbstractHumans recognize individual faces regardless of variation in the facial view. The view-tuned face neurons in the inferior temporal (IT) cortex are regarded as the neural substrate for view-invariant face recognition. This study approximated visual features encoded by these neurons as combinations of local orientations and colors, originated from natural image fragments. The resultant features reproduced the preference of these neurons to particular facial views. We also found that faces of one identity were separable from the faces of other identities in a space where each axis represented one of these features. These results suggested that view-invariant face representation was established by combining view sensitive visual features. The face representation with these features suggested that, with respect to view-invariant face representation, the seemingly complex and deeply layered ventral visual pathway can be approximated via a shallow network, comprised of layers of low-level processing for local orientations and colors (V1/V2-level) and the layers which detect particular sets of low-level elements derived from natural image fragments (IT-level).

Download Full-text

Face Perception in Face Transplant Patients

Facial Plastic Surgery ◽

10.1055/s-0038-1666786 ◽

2019 ◽

Vol 35 (05) ◽

pp. 525-533

Author(s):

Evrim Gülbetekin ◽

Seda Bayraktar ◽

Özlenen Özkan ◽

Hilmi Uysal ◽

Ömer Özkan

Keyword(s):

Face Recognition ◽

Object Recognition ◽

Face Processing ◽

Recognition Performance ◽

Discrimination Performance ◽

Object Discrimination ◽

Face Transplantation ◽

Transplant Patients ◽

Facial Surgery ◽

The Face

AbstractThe authors tested face discrimination, face recognition, object discrimination, and object recognition in two face transplantation patients (FTPs) who had facial injury since infancy, a patient who had a facial surgery due to a recent wound, and two control subjects. In Experiment 1, the authors showed them original faces and morphed forms of those faces and asked them to rate the similarity between the two. In Experiment 2, they showed old, new, and implicit faces and asked whether they recognized them or not. In Experiment 3, they showed them original objects and morphed forms of those objects and asked them to rate the similarity between the two. In Experiment 4, they showed old, new, and implicit objects and asked whether they recognized them or not. Object discrimination and object recognition performance did not differ between the FTPs and the controls. However, the face discrimination performance of FTP2 and face recognition performance of the FTP1 were poorer than that of the controls were. Therefore, the authors concluded that the structure of the face might affect face processing.

Download Full-text

Robust Image Labeling Using Conditional Random Fields

10.32920/ryerson.14651541 ◽

2021 ◽

Author(s):

Maryam Nematollahi Arani

Keyword(s):

Object Recognition ◽

Mixture Models ◽

Random Fields ◽

Conditional Random Fields ◽

Semantic Gap ◽

Visual Features ◽

Image Labeling ◽

Low Level ◽

Robust Image ◽

High Level

Object recognition has become a central topic in computer vision applications such as image search, robotics and vehicle safety systems. However, it is a challenging task due to the limited discriminative power of low-level visual features in describing the considerably diverse range of high-level visual semantics of objects. Semantic gap between low-level visual features and high-level concepts are a bottleneck in most systems. New content analysis models need to be developed to bridge the semantic gap. In this thesis, algorithms based on conditional random fields (CRF) from the class of probabilistic graphical models are developed to tackle the problem of multiclass image labeling for object recognition. Image labeling assigns a specific semantic category from a predefined set of object classes to each pixel in the image. By well capturing spatial interactions of visual concepts, CRF modeling has proved to be a successful tool for image labeling. This thesis proposes novel approaches to empowering the CRF modeling for robust image labeling. Our primary contributions are twofold. To better represent feature distributions of CRF potentials, new feature functions based on generalized Gaussian mixture models (GGMM) are designed and their efficacy is investigated. Due to its shape parameter, GGMM can provide a proper fit to multi-modal and skewed distribution of data in nature images. The new model proves more successful than Gaussian and Laplacian mixture models. It also outperforms a deep neural network model on Corel imageset by 1% accuracy. Further in this thesis, we apply scene level contextual information to integrate global visual semantics of the image with pixel-wise dense inference of fully-connected CRF to preserve small objects of foreground classes and to make dense inference robust to initial misclassifications of the unary classifier. Proposed inference algorithm factorizes the joint probability of labeling configuration and image scene type to obtain prediction update equations for labeling individual image pixels and also the overall scene type of the image. The proposed context-based dense CRF model outperforms conventional dense CRF model by about 2% in terms of labeling accuracy on MSRC imageset and by 4% on SIFT Flow imageset. Also, the proposed model obtains the highest scene classification rate of 86% on MSRC dataset.

Download Full-text

Global Context Extraction for Object Recognition Using a Combination of Range and Visual Features

Dynamic 3D Imaging - Lecture Notes in Computer Science ◽

10.1007/978-3-642-03778-8_8 ◽

2009 ◽

pp. 96-109 ◽

Cited By ~ 2

Author(s):

Michael Kemmler ◽

Erik Rodner ◽

Joachim Denzler

Keyword(s):

Object Recognition ◽

Visual Features ◽

Global Context ◽

Context Extraction

Download Full-text

Face Recognition Challenge: Object Recognition Approaches for Human/Avatar Classification

2012 11th International Conference on Machine Learning and Applications ◽

10.1109/icmla.2012.188 ◽

2012 ◽

Author(s):

Toshihiko Yamasaki ◽

Tsuhan Chen

Keyword(s):

Face Recognition ◽

Object Recognition

Download Full-text

An Exploratory Study of the Relationship between Face Recognition Memory and the Volume of Medial Temporal Lobe Structures in Healthy Young Males

Behavioural Neurology ◽

10.1155/1998/285061 ◽

1998 ◽

Vol 11 (1) ◽

pp. 3-20 ◽

Cited By ~ 30

Author(s):

Clare E. Mackay ◽

Neil Roberts ◽

Andrew R. Mayes ◽

John J. Downes ◽

Jonathan K. Foster ◽

...

Keyword(s):

Face Recognition ◽

Recognition Memory ◽

Temporal Lobe ◽

Medial Temporal Lobe ◽

Multiple Comparisons ◽

Cerebral Hemispheres ◽

Parahippocampal Gyrus ◽

Bonferroni Correction ◽

Temporal Pole ◽

The Right

A rigorous new methodology was applied to the study of structure function relationships in the living human brain. Face recognition memory (FRM) and other cognitive measures were made in 29 healthy young male subjects (mean age = 21.7 years) and related to volumetric measurements of their cerebral hemispheres and of structures in their medial temporal lobes, obtained using the Cavalieri method in combination with high resolution Magnetic Resonance Imaging (MRI. Greatest proportional variability in volumes was found for the lateral ventricles (57%) for the cerebral hemispheres (8%) in the mean volumes of the hippocampus, parahippocampal gyrus, amygdala, caudate nucleus, temporal pole and temporal lobe on the right and left sides of the brain. The volumes of the right and left parahippocampal gyrus, temporal pole, temporal lobe, and left hippocampus were, prior to application of the Bonferroni correction to take account of 12 multiple comparisons, significantly correlated with the volume of the corresponding hemisphere (p< 0.05). The volumes of all structures were highly correlated (p< 0.0002 for all comparisons) between the two cerebral hemispheres. There were no positive relationships between structure volumes and FRM score. However, the volume of the right amygdala was, prior to application of the Bonferroni correction to take account of 38~multiple comparisons, found to be significantly smaller in the five most consistent high scorers compared to the five most consistent low scorers (t= 2.77,p= 0.025). The implications for possible relationships between healthy medial temporal lobe structures and memory are discussed.

Download Full-text

Face Recognition Is Affected by Similarity in Spatial Frequency Range to a Greater Degree Than Within-Category Object Recognition.

Journal of Experimental Psychology Human Perception & Performance ◽

10.1037/0096-1523.30.5.975 ◽

2004 ◽

Vol 30 (5) ◽

pp. 975-987 ◽

Cited By ~ 19

Author(s):

Charles A. Collin ◽

Chang Hong Liu ◽

Nikolaus F. Troje ◽

Patricia A. McMullen ◽

Avi Chaudhuri

Keyword(s):

Face Recognition ◽

Object Recognition ◽

Spatial Frequency ◽

Frequency Range

Download Full-text

Hillary Clinton as Campaign Surrogate

10.5406/illinois/9780252038211.003.0002 ◽

2017 ◽

Author(s):

Shawn J. Parry-Giles

Keyword(s):

News Coverage ◽

Presidential Campaign ◽

Hillary Clinton ◽

News Frames ◽

First Lady ◽

Presidential Campaigning

This chapter represents the first installment of Hillary Clinton's news biography and examines the news coverage of Clinton during the 1992 presidential campaign and her entrance onto the national stage of politics. It recounts the baseline news frames that laid the foundation for judgments of Clinton's authenticity against which future frames would converge and diverge. The chapter also describes her most formative media moments during this period, which linguistically and visually acted as stock frames that authenticated Clinton as a feminist and inauthenticated her as a woman of tradition. Her political image was thus framed as a political intruder violating the protocol of presidential campaigning; an anomalous candidate's wife rejecting the trappings of home and domesticity in favor of feminist principles; and a political lightning rod who exuded personality problems that promised to disrupt her husband's presidential bid and undermine the traditions of first lady.

Download Full-text