Principles Of Predictive Coding Research Articles

Active Vision in Binocular Depth Estimation: A Top-Down Perspective.

Depth estimation is an ill-posed problem; objects of different shapes or dimensions, even if at different distances, may project to the same image on the retina. Our brain uses several cues for depth estimation, including monocular cues such as motion parallax and binocular cues such as diplopia. However, it remains unclear how the computations required for depth estimation are implemented in biologically plausible ways. State-of-the-art approaches to depth estimation based on deep neural networks implicitly describe the brain as a hierarchical feature detector. Instead, in this paper we propose an alternative approach that casts depth estimation as a problem of active inference. We show that depth can be inferred by inverting a hierarchical generative model that simultaneously predicts the eyes' projections from a 2D belief over an object. Model inversion consists of a series of biologically plausible homogeneous transformations based on Predictive Coding principles. Under the plausible assumption of a nonuniform fovea resolution, depth estimation favors an active vision strategy that fixates the object with the eyes, rendering the depth belief more accurate. This strategy is not realized by first fixating on a target and then estimating the depth; instead, it combines the two processes through action-perception cycles, with a similar mechanism of the saccades during object recognition. The proposed approach requires only local (top-down and bottom-up) message passing, which can be implemented in biologically plausible neural circuits.

Modeling face recognition in the predictive coding framework: A combined computational modeling and functional imaging study

The learning of new facial identities and the recognition of familiar faces are crucial processes for social interactions. Recently, a combined computational modeling and functional magnetic resonance imaging (fMRI) study used predictive coding as a biologically plausible framework to model face identity learning and to relate specific model parameters with brain activity (Apps and Tsakiris, Nat Commun 4, 2698, 2013). On the one hand, it was shown that behavioral responses on a two-option face recognition task could be predicted by the level of contextual and facial familiarity in a computational model derived from predictive-coding principles. On the other hand, brain activity in specific brain regions was associated with these parameters. More specifically, brain activity in the superior temporal sulcus (STS) varied with contextual familiarity, whereas activity in the fusiform face area (FFA) covaried with the prediction error parameter that updated facial familiarity.Literature combining fMRI assessments and computational modeling in humans still needs to be expanded. Furthermore, prior results are largely not replicated. The present study was, therefore, specifically set up to replicate these previous findings. Our results support the original findings in two critical aspects. First, on a group level, the behavioral responses were modeled best by the same computational model reported by the original authors. Second, we showed that estimates of these model parameters covary with brain activity in specific, face-sensitive brain regions. Our results thus provide further evidence that the functional properties of the face perception network conform to central principles of predictive coding. However, our study yielded diverging findings on specific computational model parameters reflected in brain activity. On the one hand, we did not find any evidence of a computational involvement of the STS. On the other hand, our results showed that activity in the right FFA was associated with multiple computational model parameters. Our data do not provide evidence for functional segregation between particular face-sensitive brain regions, as previously proposed.

Inconsistent illusory motion in predictive coding deep neural networks

Why do we perceive illusory motion in some static images? Several accounts point to eye movements, response latencies to different image elements, or interactions between image patterns and motion energy detectors. Recently PredNet, a recurrent deep neural network (DNN) based on predictive coding principles, was reported to reproduce the “Rotating Snakes” illusion, suggesting a role for predictive coding. We begin by replicating this finding, then use a series of “in silico” psychophysics and electrophysiology experiments to examine whether PredNet behaves consistently with human observers and non-human primate neural data. A pretrained PredNet predicted illusory motion for all subcomponents of the Rotating Snakes pattern, consistent with human observers. However, we found no simple response delays in internal units, unlike evidence from electrophysiological data. PredNet’s detection of motion in gradients seemed dependent on contrast, but depends predominantly on luminance in humans. Finally, we examined the robustness of the illusion across ten PredNets of identical architecture, retrained on the same video data. There was large variation across network instances in whether they reproduced the Rotating Snakes illusion, and what motion, if any, they predicted for simplified variants. Unlike human observers, no network predicted motion for greyscale variants of the Rotating Snakes pattern. Our results sound a cautionary note: even when a DNN successfully reproduces some idiosyncrasy of human vision, more detailed investigation can reveal inconsistencies between humans and the network, and between different instances of the same network. These inconsistencies suggest that predictive coding does not reliably give rise to human-like illusory motion.

A Predictive Coding Framework for Understanding Major Depression.

Predictive coding models of brain processing propose that top-down cortical signals promote efficient neural signaling by carrying predictions about incoming sensory information. These “priors” serve to constrain bottom-up signal propagation where prediction errors are carried via feedforward mechanisms. Depression, traditionally viewed as a disorder characterized by negative cognitive biases, is associated with disrupted reward prediction error encoding and signaling. Accumulating evidence also suggests that depression is characterized by impaired local and long-range prediction signaling across multiple sensory domains. This review highlights the electrophysiological and neuroimaging evidence for disrupted predictive processing in depression. The discussion is framed around the manner in which disrupted generative predictions about the sensorium could lead to depressive symptomatology, including anhedonia and negative bias. In particular, the review focuses on studies of sensory deviance detection and reward processing, highlighting research evidence for both disrupted generative predictions and prediction error signaling in depression. The role of the monoaminergic and glutamatergic systems in predictive coding processes is also discussed. This review provides a novel framework for understanding depression using predictive coding principles and establishes a foundational roadmap for potential future research.

Neurons learn by predicting future activity

Understanding how the brain learns may lead to machines with human-like intellectual capacities. It was previously proposed that the brain may operate on the principle of predictive coding. However, it is still not well understood how a predictive system could be implemented in the brain. Here we demonstrate that the ability of a single neuron to predict its future activity may provide an effective learning mechanism. Interestingly, this predictive learning rule can be derived from a metabolic principle, whereby neurons need to minimize their own synaptic activity (cost) while maximizing their impact on local blood supply by recruiting other neurons. We show how this mathematically derived learning rule can provide a theoretical connection between diverse types of brain-inspired algorithm, thus offering a step towards the development of a general theory of neuronal learning. We tested this predictive learning rule in neural network simulations and in data recorded from awake animals. Our results also suggest that spontaneous brain activity provides ‘training data’ for neurons to learn to predict cortical dynamics. Thus, the ability of a single neuron to minimize surprise—that is, the difference between actual and expected activity—could be an important missing element to understand computation in the brain.

The Potential Role of Dopamine in Mediating Motor Function and Interpersonal Synchrony

Motor functions in general and motor planning in particular are crucial for our ability to synchronize our movements with those of others. To date, these co-occurring functions have been studied separately, and as yet it is unclear whether they share a common biological mechanism. Here, we synthesize disparate recent findings on motor functioning and interpersonal synchrony and propose that these two functions share a common neurobiological mechanism and adhere to the same principles of predictive coding. Critically, we describe the pivotal role of the dopaminergic system in modulating these two distinct functions. We present attention deficit hyperactivity disorder (ADHD) as an example of a disorder that involves the dopaminergic system and describe deficits in motor and interpersonal synchrony. Finally, we suggest possible directions for future studies emphasizing the role of dopamine modulation as a link between social and motor functioning.

Regulation of Functions of the Brain and Body by the Principle of predictive Coding

Lisa Feldman-Barrett, que promovió la teoría constructivista del afecto, recientemente ha propuesto el modelo de afecto la Codificación predictiva incorporada de la interocepción (EPIC), basado en la perspectiva de la codificación predictiva. El marco teórico de la codificación predictiva discute los hechos de que el cerebro crea modelos internos que puedan facilitar predicciones para la percepción y las habilidades motoras, y de que la percepción y la conducta surgen de los cálculos bayesianos que se basan en estas predicciones. El modelo EPIC expande este marco a la interocepción – la percepción del cuerpo interno, y trata de explicar el fenómeno de afecto como experiencia integrativa basada en la interocepción. Esta perspectiva ofrece implicaciones importantes para entender los problemas del eje cerebro-intestino y sus discapacidades.

Shamanism and psychosis: Shared mechanisms?

Individual-specific predispositions may precede the cultural evolution of shamanism and may be linked to it via principles of predictive coding. We have used these principles to identify commonalities between clinical and shaman-like non-clinical voice-hearers. The author may find this approach helpful in relating the experiences of shamans to those of their clients.

Investigating bottom-up auditory attention

In everyday life, we are surrounded by a cacophony of sounds that our brain has to sort through in order to focus on important information. The conspicuity of certain sounds allows them to grab our attention and “stand out” relative to the soundscape. In this work, we hypothesize that as sound statistics change over time, variance from these statistics drive bottom-up attention processes that direct our focus to certain objects in the scene. We investigate the perceptual space that renders certain sounds salient relative to their context, and examine neural underpinnings of emergence of these salient sounds using complex acoustic scenes as stimuli. Results reveal adaptive neural representations of the context and salient objects reflecting deviations from the statistical structure of the scene. We speculate that principles of predictive coding could explain a number of observed neural and perceptual findings and discuss relevance of such framework in understanding processes of auditory scene analysis.

Predictive codes of familiarity and context during the perceptual learning of facial identities

Face recognition is a key component of successful social behaviour. However, the computational processes that underpin perceptual learning and recognition as faces transition from unfamiliar to familiar are poorly understood. In predictive coding, learning occurs through prediction errors that update stimulus familiarity, but recognition is a function of both stimulus and contextual familiarity. Here we show that behavioural responses on a two-option face recognition task can be predicted by the level of contextual and facial familiarity in a computational model derived from predictive-coding principles. Using fMRI, we show that activity in the superior temporal sulcus varies with the contextual familiarity in the model, whereas activity in the fusiform face area covaries with the prediction error parameter that updated facial familiarity. Our results characterize the key computations underpinning the perceptual learning of faces, highlighting that the functional properties of face-processing areas conform to the principles of predictive coding.

The resource model and the principle of predictive coding: a framework for analyzing proximate effects of ritual

The resource model and the principle of predictive coding: a framework for analyzing proximate effects of ritual

Developing a Diagnostic Brain Imaging Protocol for Recent Onset Psychosis which Embodies Principles of Predictive Coding and Free Energy

Although schizophrenia is usually conceptualized of as a disorder of well formed hallucinations and delusions, its onset is more typically characterized by heightened arousal, a vague but pervasive feeling of unease and poorly formed misperceptions. This unstable constellation of symptoms challenges clinical diagnosis at a time when properly guided interventions are most likely to have the greatest long-term benefits. It is possible to frame this clinical picture as a disturbance in the capacity of the cortex to optimally form predictive models of the environment, to appropriately sample visual scenes in order to estimate the likelihood of such models and to hence minimize surprise in a dynamic social landscape. This approach suggests that manipulating the relationship between visual search strategies and natural scene statistics might be a sensitive means of quantifying core neurobiological deficits early in the course of emerging psychotic disorders. We are developing a diagnostic tool for the prodromal phase of these disorders which involves experimentally disturbing the relationship between visual saccades and the stream of natural images in well directed films. I will outline the conceptual and computational bases and early experimental results thus far obtained, including a canonical example of paranoia and theory of mind in a spaghetti western.

Predictive Coding and Pitch Processing in the Auditory Cortex

In this work, we show that electrophysiological responses during pitch perception are best explained by distributed activity in a hierarchy of cortical sources and, crucially, that the effective connectivity between these sources is modulated with pitch strength. Local field potentials were recorded in two subjects from primary auditory cortex and adjacent auditory cortical areas along the axis of Heschl's gyrus (HG) while they listened to stimuli of varying pitch strength. Dynamic causal modeling was used to compare system architectures that might explain the recorded activity. The data show that representation of pitch requires an interaction between nonprimary and primary auditory cortex along HG that is consistent with the principle of predictive coding.

Predictive Coding and the Slowness Principle: An Information-Theoretic Approach

Understanding the guiding principles of sensory coding strategies is a main goal in computational neuroscience. Among others, the principles of predictive coding and slowness appear to capture aspects of sensory processing. Predictive coding postulates that sensory systems are adapted to the structure of their input signals such that information about future inputs is encoded. Slow feature analysis (SFA) is a method for extracting slowly varying components from quickly varying input signals, thereby learning temporally invariant features. Here, we use the information bottleneck method to state an information-theoretic objective function for temporally local predictive coding. We then show that the linear case of SFA can be interpreted as a variant of predictive coding that maximizes the mutual information between the current output of the system and the input signal in the next time step. This demonstrates that the slowness principle and predictive coding are intimately related.