Joint encoding of facial identity, orientation, gaze, and expression in the middle dorsal face area

The state of modelling face processing in humans with deep learning

Deep learning models trained for facial recognition now surpass the highest performing human participants. Recent evidence suggests that they also model some qualitative aspects of face processing in humans. This review compares the current understanding of deep learning models with psychological models of the face processing system. Psychological models consist of two components that operate on the information encoded when people perceive a face, which we refer to here as 'face codes'. The first component, the core system, extracts face codes from retinal input that encode invariant and changeable properties. The second component, the extended system, links face codes to personal information about a person and their social context. Studies of face codes in existing deep learning models reveal some surprising results. For example, face codes in networks designed for identity recognition also encode expression information, which contrasts with psychological models that separate invariant and changeable properties. Deep learning can also be used to implement candidate models of the face processing system, for example to compare alternative cognitive architectures and codes that might support interchange between core and extended face processing systems. We conclude by summarizing seven key lessons from this research and outlining three open questions for future study.

Facial gestures are enacted via a cortical hierarchy of dynamic and stable codes

Successful communication requires the generation and perception of a shared set of signals. Facial gestures are one fundamental set of communicative behaviors in primates, generated through the dynamic arrangement of dozens of fine muscles. While much progress has been made uncovering the neural mechanisms of face perception, little is known about those controlling facial gesture production. Commensurate with the importance of facial gestures in daily social life, anatomical work has shown that facial muscles are under direct control from multiple cortical regions, including primary and premotor in lateral frontal cortex, and cingulate in medial frontal cortex. Furthermore, neuropsychological evidence from focal lesion patients has suggested that lateral cortex controls voluntary movements, and medial emotional expressions. Here we show that lateral and medial cortical face motor regions encode both types of gestures. They do so through unique temporal activity patterns, distinguishable well-prior to movement onset. During gesture production, cortical regions encoded facial kinematics in a context-dependent manner. Our results show how cortical regions projecting in parallel downstream, but each situated at a different level of a posterior-anterior hierarchy form a continuum of gesture coding from dynamic to temporally stable, in order to produce context-related, coherent motor outputs during social communication.

Neuronal Mechanisms Underlying Face Recognition in Non-human Primates

Humans and primates rely on visual face recognition for social interactions. Damage to specific brain areas causes prosopagnosia, a condition characterized by the inability to recognize familiar faces, indicating the presence of specialized brain areas for face processing. A breakthrough finding came from a non-human primate (NHP) study conducted in the early 2000s; it was the first to identify multiple face processing areas in the temporal lobe, termed face patches. Subsequent studies have demonstrated the unique role of each face patch in the structural analysis of faces. More recent studies have expanded these findings by exploring the role of face patch networks in social and memory functions and the importance of early face exposure in the development of the system. In this review, we discuss the neuronal mechanisms responsible for analyzing facial features, categorizing faces, and associating faces with memory and social contexts within both the cerebral cortex and subcortical areas. Use of NHPs in neuropsychological and neurophysiological studies can highlight the mechanistic aspects of the neuronal circuit underlying face recognition at both the single-neuron and whole-brain network levels.

Human Recognition: The Utilization of Face, Voice, Name and Interactions-An Extended Editorial

The many stimulating contributions to this Special Issue of Brain Science focused on some basic issues of particular interest in current research, with emphasis on human recognition using faces, voices, and names [...].

Neuronal Population Encoding of Identity in Primate Prefrontal Cortex

The ventrolateral prefrontal cortex (VLPFC) shows robust activation during the perception of faces and voices. However, little is known about what categorical features of social stimuli drive neural activity in this region. Since perception of identity and expression are critical social functions, we examined whether neural responses to naturalistic stimuli were driven by these two categorical features in the prefrontal cortex. We recorded single neurons in the VLPFC, while two male rhesus macaques (Macaca mulatta) viewed short audiovisual videos of unfamiliar conspecifics making expressions of aggressive, affiliative, and neutral valence. Of the 285 neurons responsive to the audiovisual stimuli, 111 neurons had a main effect (two-way ANOVA) of identity, expression, or their interaction in their stimulus-related firing rates; however, decoding of expression and identity using single-unit firing rates rendered poor accuracy. Interestingly, when decoding from pseudo-populations of recorded neurons, the accuracy for both expression and identity increased with population size, suggesting that the population transmitted information relevant to both variables. Principal components analysis of mean population activity across time revealed that population responses to the same identity followed similar trajectories in the response space, facilitating segregation from other identities. Our results suggest that identity is a critical feature of social stimuli that dictates the structure of population activity in the VLPFC, during the perception of vocalizations and their corresponding facial expressions. These findings enhance our understanding of the role of the VLPFC in social behavior.

Bodies in motion: Unraveling the distinct roles of motion and shape in dynamic body responses in the temporal cortex

The temporal cortex represents social stimuli, including bodies. We examine and compare the contributions of dynamic and static features to the single-unit responses to moving monkey bodies in and between a patch in the anterior dorsal bank of the superior temporal sulcus (dorsal patch [DP]) and patches in the anterior inferotemporal cortex (ventral patch [VP]), using fMRI guidance in macaques. The response to dynamics varies within both regions, being higher in DP. The dynamic body selectivity of VP neurons correlates with static features derived from convolutional neural networks and motion. DP neurons' dynamic body selectivity is not predicted by static features but is dominated by motion. Whereas these data support the dominance of motion in the newly proposed "dynamic social perception" stream, they challenge the traditional view that distinguishes DP and VP processing in terms of motion versus static features, underscoring the role of inferotemporal neurons in representing body dynamics.

Multisensory interactions of face and vocal information during perception and memory in ventrolateral prefrontal cortex

The ventral frontal lobe is a critical node in the circuit that underlies communication, a multisensory process where sensory features of faces and vocalizations come together. The neural basis of face and vocal integration is a topic of great importance since the integration of multiple sensory signals is essential for the decisions that govern our social interactions. Investigations have shown that the macaque ventrolateral prefrontal cortex (VLPFC), a proposed homologue of the human inferior frontal gyrus, is involved in the processing, integration and remembering of audiovisual signals. Single neurons in VLPFC encode and integrate species-specific faces and corresponding vocalizations. During working memory, VLPFC neurons maintain face and vocal information online and exhibit selective activity for face and vocal stimuli. Population analyses indicate that identity, a critical feature of social stimuli, is encoded by VLPFC neurons and dictates the structure of dynamic population activity in the VLPFC during the perception of vocalizations and their corresponding facial expressions. These studies suggest that VLPFC may play a primary role in integrating face and vocal stimuli with contextual information, in order to support decision making during social communication. This article is part of the theme issue 'Decision and control processes in multisensory perception'.

Causal Manipulation of Gaze-Following in the Macaque Temporal Cortex

Gaze-following, the ability to shift one's own attention to places or objects others are looking at, is essential for social interactions. Single unit recordings from the monkey cortex and neuroimaging work on the human and monkey brain suggest that a distinct region in the temporal cortex, the gaze-following patch (GFP), underpins this ability. Since previous studies of the GFP have relied on correlational techniques, it remains unclear whether gaze-following related activity in the GFP indicates a causal role rather than being just a reverberation of behaviorally relevant information produced elsewhere. To answer this question, we applied focal electrical and pharmacological perturbation to the GFP. Both approaches, when applied to the GFP, disrupted gaze-following if the monkeys had been instructed to follow gaze, along with the ability to suppress it if vetoed by the context. Hence the GFP is necessary for gaze-following as well as its cognitive control.

The contribution of dynamics to macaque body and face patch responses

Previous functional imaging studies demonstrated body-selective patches in the primate visual temporal cortex, comparing activations to static bodies and static images of other categories. However, the use of static instead of dynamic displays of moving bodies may have underestimated the extent of the body patch network. Indeed, body dynamics provide information about action and emotion and may be processed in patches not activated by static images. Thus, to map with fMRI the full extent of the macaque body patch system in the visual temporal cortex, we employed dynamic displays of natural-acting monkey bodies, dynamic monkey faces, objects, and scrambled versions of these videos, all presented during fixation. We found nine body patches in the visual temporal cortex, starting posteriorly in the superior temporal sulcus (STS) and ending anteriorly in the temporal pole. Unlike for static images, body patches were present consistently in both the lower and upper banks of the STS. Overall, body patches showed a higher activation by dynamic displays than by matched static images, which, for identical stimulus displays, was less the case for the neighboring face patches. These data provide the groundwork for future single-unit recording studies to reveal the spatiotemporal features the neurons of these body patches encode. These fMRI findings suggest that dynamics have a stronger contribution to population responses in body than face patches.

Research on social attention in autism and the challenges of the research domain criteria (RDoC) framework

The fuzzy nature of categories of psychopathology, such as autism, leads to significant research challenges. Alternatively, focusing research on the study of a common set of important and well-defined psychological constructs across psychiatric conditions may make the fundamental etiological processes of psychopathology easier to discern and treat (Cuthbert, 2022). The development of the research domain criteria (RDoC) framework is designed to guide this new research approach (Insel et al., 2010). However, progress in research may be expected to continually refine and reorganize the understanding of the specifics of these mental processes (Cuthbert & Insel, 2013). Moreover, knowledge gleaned from the study of both normative and atypical development can be mutually informative in the evolution of our understanding of these fundamental processes. A case in point is the study of social attention. This Autism 101 commentary provides an educational summary of research over the last few decades indicates that social attention is major construct in the study of human social-cognitive development, autism and other forms of psychopathology. The commentary also describes how this research can inform the Social Process dimension of the RDoC framework.

Encoding of dynamic facial information in the middle dorsal face area

Faces in motion reveal a plethora of information through visual dynamics. Faces can move in complex patterns while transforming facial shape, e.g., during the generation of different emotional expressions. While motion and shape processing have been studied extensively in separate research enterprises, much less is known about their conjunction during biological motion. Here, we took advantage of the discovery in brain-imaging studies of an area in the dorsal portion of the macaque monkey superior temporal sulcus (STS), the middle dorsal face area (MD), with selectivity for naturalistic face motion. To gain mechanistic insights into the coding of facial motion, we recorded single-unit activity from MD, testing whether and how MD cells encode face motion. The MD population was highly sensitive to naturalistic facial motion and facial shape. Some MD cells responded only to the conjunction of facial shape and motion, others were selective for facial shape even without movement, and yet others were suppressed by facial motion. We found that this heterogeneous MD population transforms face motion into a higher dimensional activity space, a representation that would allow for high sensitivity to relevant small-scale movements. Indeed, we show that many MD cells carry such sensitivity for eye movements. We further found that MD cells encode motion of head, mouth, and eyes in a separable manner, requiring the use of multiple reference frames. Thus, MD is a bona fide face-motion area that uses highly heterogeneous cell populations to create codes capturing even complex facial motion trajectories.

Challenging the Classical View: Recognition of Identity and Expression as Integrated Processes

Recent neuroimaging evidence challenges the classical view that face identity and facial expression are processed by segregated neural pathways, showing that information about identity and expression are encoded within common brain regions. This article tests the hypothesis that integrated representations of identity and expression arise spontaneously within deep neural networks. A subset of the CelebA dataset is used to train a deep convolutional neural network (DCNN) to label face identity (chance = 0.06%, accuracy = 26.5%), and the FER2013 dataset is used to train a DCNN to label facial expression (chance = 14.2%, accuracy = 63.5%). The identity-trained and expression-trained networks each successfully transfer to labeling both face identity and facial expression on the Karolinska Directed Emotional Faces dataset. This study demonstrates that DCNNs trained to recognize face identity and DCNNs trained to recognize facial expression spontaneously develop representations of facial expression and face identity, respectively. Furthermore, a congruence coefficient analysis reveals that features distinguishing between identities and features distinguishing between expressions become increasingly orthogonal from layer to layer, suggesting that deep neural networks disentangle representational subspaces corresponding to different sources.

Representational structure of fMRI/EEG responses to dynamic facial expressions

Face perception provides an excellent example of how the brain processes nuanced visual differences and transforms them into behaviourally useful representations of identities and emotional expressions. While a body of literature has looked into the spatial and temporal neural processing of facial expressions, few studies have used a dimensionally varying set of stimuli containing subtle perceptual changes. In the current study, we used 48 short videos varying dimensionally in their intensity and category (happy, angry, surprised) of expression. We measured both fMRI and EEG responses to these video clips and compared the neural response patterns to the predictions of models based on image features and models derived from behavioural ratings of the stimuli. In fMRI, the inferior frontal gyrus face area (IFG-FA) carried information related only to the intensity of the expression, independent of image-based models. The superior temporal sulcus (STS), inferior temporal (IT) and lateral occipital (LO) areas contained information about both expression category and intensity. In the EEG, the coding of expression category and low-level image features were most pronounced at around 400 ms. The expression intensity model did not, however, correlate significantly at any EEG timepoint. Our results show a specific role for IFG-FA in the coding of expressions and suggest that it contains image and category invariant representations of expression intensity.

Joint encoding of facial identity, orientation, gaze, and expression in the middle dorsal face area

The last two decades have established that a network of face-selective areas in the temporal lobe of macaque monkeys supports the visual processing of faces. Each area within the network contains a large fraction of face-selective cells. And each area encodes facial identity and head orientation differently. A recent brain-imaging study discovered an area outside of this network selective for naturalistic facial motion, the middle dorsal (MD) face area. This finding offers the opportunity to determine whether coding principles revealed inside the core network would generalize to face areas outside the core network. We investigated the encoding of static faces and objects, facial identity, and head orientation, dimensions which had been studied in multiple areas of the core face-processing network before, as well as facial expressions and gaze. We found that MD populations form a face-selective cluster with a degree of selectivity comparable to that of areas in the core face-processing network. MD encodes facial identity robustly across changes in head orientation and expression, it encodes head orientation robustly against changes in identity and expression, and it encodes expression robustly across changes in identity and head orientation. These three dimensions are encoded in a separable manner. Furthermore, MD also encodes the direction of gaze in addition to head orientation. Thus, MD encodes both structural properties (identity) and changeable ones (expression and gaze) and thus provides information about another animal's direction of attention (head orientation and gaze). MD contains a heterogeneous population of cells that establish a multidimensional code for faces.