Bayesian Computation in Recurrent Neural Circuits

2004 ◽  
Vol 16 (1) ◽  
pp. 1-38 ◽  
Author(s):  
Rajesh P. N. Rao
Keyword(s):  
Motion Detection ◽  
Network Architecture ◽  
Discrimination Task ◽  
Visual Motion ◽  
Natural World ◽  
Direction Selectivity ◽  
Orientation Discrimination ◽  
Posterior Probabilities ◽  
Motion Direction ◽  
Model Network

A large number of human psychophysical results have been successfully explained in recent years using Bayesian models. However, the neural implementation of such models remains largely unclear. In this article, we show that a network architecture commonly used to model the cerebral cortex can implement Bayesian inference for an arbitrary hidden Markov model. We illustrate the approach using an orientation discrimination task and a visual motion detection task. In the case of orientation discrimination, we show that the model network can infer the posterior distribution over orientations and correctly estimate stimulus orientation in the presence of significant noise. In the case of motion detection, we show that the resulting model network exhibits direction selectivity and correctly computes the posterior probabilities over motion direction and position. When used to solve the well-known random dots motion discrimination task, the model generates responses that mimic the activities of evidence-accumulating neurons in cortical areas LIP and FEF. The framework we introduce posits a new interpretation of cortical activities in terms of log posterior probabilities of stimuli occurring in the natural world.

scholarly journals Neural selectivity for visual motion in macaque area V3A

2020 ◽  
Author(s):  
Nardin Nakhla ◽  
Yavar Korkian ◽  
Matthew R. Krause ◽  
Christopher C. Pack
Keyword(s):  
Visual Cortex ◽  
Optic Flow ◽  
Functional Role ◽  
Visual Motion ◽  
Flow Patterns ◽  
Brain Regions ◽  
Direction Selectivity ◽  
Motion Processing ◽  
Imaging Studies ◽  
Motion Direction

AbstractThe processing of visual motion is carried out by dedicated pathways in the primate brain. These pathways originate with populations of direction-selective neurons in the primary visual cortex, which project to dorsal structures like the middle temporal (MT) and medial superior temporal (MST) areas. Anatomical and imaging studies have suggested that area V3A might also be specialized for motion processing, but there have been very few studies of single-neuron direction selectivity in this area. We have therefore performed electrophysiological recordings from V3A neurons in two macaque monkeys (one male and one female) and measured responses to a large battery of motion stimuli that includes translation motion, as well as more complex optic flow patterns. For comparison, we simultaneously recorded the responses of MT neurons to the same stimuli. Surprisingly, we find that overall levels of direction selectivity are similar in V3A and MT and moreover that the population of V3A neurons exhibits somewhat greater selectivity for optic flow patterns. These results suggest that V3A should be considered as part of the motion processing machinery of the visual cortex, in both human and non-human primates.Significance statementAlthough area V3A is frequently the target of anatomy and imaging studies, little is known about its functional role in processing visual stimuli. Its contribution to motion processing has been particularly unclear, with different studies yielding different conclusions. We report a detailed study of direction selectivity in V3A. Our results show that single V3A neurons are, on average, as capable of representing motion direction as are neurons in well-known structures like MT. Moreover, we identify a possible specialization for V3A neurons in representing complex optic flow, which has previously been thought to emerge in higher-order brain regions. Thus it appears that V3A is well-suited to a functional role in motion processing.

Retinal Mechanisms for Motion Detection

2021 ◽  
Author(s):  
Mathew T. Summers ◽  
Malak El Quessny ◽  
Maria B. Feller
Keyword(s):  
Motion Detection ◽  
Visual Motion ◽  
Sensory Experience ◽  
Direction Selectivity ◽  
Model System ◽  
Neural Computation ◽  
Motion Sensors ◽  
Neural Signals ◽  
Mammalian Retina ◽  
Detailed Circuit

Motion is a key feature of the sensory experience of visual animals. The mammalian retina has evolved a number of diverse motion sensors to detect and parse visual motion into behaviorally relevant neural signals. Extensive work has identified retinal outputs encoding directional and nondirectional motion, and the intermediate circuitry underlying this tuning. Detailed circuit mechanism investigation has established retinal direction selectivity in particular as a model system of neural computation.

Linking Neuronal Direction Selectivity to Perceptual Decisions About Visual Motion

2020 ◽  
Vol 6 (1) ◽  
pp. 335-362
Author(s):  
Tatiana Pasternak ◽  
Duje Tadin
Keyword(s):  
Prefrontal Cortex ◽  
Visual Processing ◽  
Visual Motion ◽  
Neuronal Response ◽  
Physiological Responses ◽  
Direction Selectivity ◽  
Motion Information ◽  
Motion Direction ◽  
Neurophysiological Studies ◽  
Selective Activity

Psychophysical and neurophysiological studies of responses to visual motion have converged on a consistent set of general principles that characterize visual processing of motion information. Both types of approaches have shown that the direction and speed of target motion are among the most important encoded stimulus properties, revealing many parallels between psychophysical and physiological responses to motion. Motivated by these parallels, this review focuses largely on more direct links between the key feature of the neuronal response to motion, direction selectivity, and its utilization in memory-guided perceptual decisions. These links were established during neuronal recordings in monkeys performing direction discriminations, but also by examining perceptual effects of widespread elimination of cortical direction selectivity produced by motion deprivation during development. Other approaches, such as microstimulation and lesions, have documented the importance of direction-selective activity in the areas that are active during memory-guided direction comparisons, area MT and the prefrontal cortex, revealing their likely interactions during behavioral tasks.

scholarly journals Complementary mechanisms create direction selectivity in the fly

eLife ◽  
2016 ◽  
Vol 5 ◽  
Author(s):  
Juergen Haag ◽  
Alexander Arenz ◽  
Etienne Serbe ◽  
Fabrizio Gabbiani ◽  
Alexander Borst
Keyword(s):  
Motion Detection ◽  
Visual Motion ◽  
Optic Lobe ◽  
Direction Selectivity ◽  
Neural Computation ◽  
Directional Selectivity ◽  
Motion Sensing ◽  
Null Direction ◽  
Preferred Direction ◽  
High Degree

How neurons become sensitive to the direction of visual motion represents a classic example of neural computation. Two alternative mechanisms have been discussed in the literature so far: preferred direction enhancement, by which responses are amplified when stimuli move along the preferred direction of the cell, and null direction suppression, where one signal inhibits the response to the subsequent one when stimuli move along the opposite, i.e. null direction. Along the processing chain in the Drosophila optic lobe, directional responses first appear in T4 and T5 cells. Visually stimulating sequences of individual columns in the optic lobe with a telescope while recording from single T4 neurons, we find both mechanisms at work implemented in different sub-regions of the receptive field. This finding explains the high degree of directional selectivity found already in the fly’s primary motion-sensing neurons and marks an important step in our understanding of elementary motion detection.

The Effect of Temporal Phase on the Perception of Apparent Motion

Perception ◽  
1996 ◽  
Vol 25 (1_suppl) ◽  
pp. 68-68
Author(s):  
H S Hock ◽  
K Kogan ◽  
N Lodes
Keyword(s):  
Apparent Motion ◽  
Motion Detection ◽  
Discrimination Task ◽  
Luminance Contrast ◽  
Background Luminance ◽  
Motion Direction ◽  
Direction Discrimination ◽  
Temporal Phase ◽  
The Difference ◽  
Perceived Speed

In classical apparent motion, a spot of light is presented in alternation such that the waveforms describing the varying luminance at each of two locations are 180° out of phase. However, when the luminance variation at each location is approximately sinusoidal, and the perceiver's task is to discriminate motion direction, the optimum temporal phase is 90° (van Santen and Sperling, 1984 Journal of the Optical Society of America A1 451 – 473). The results reported in this study suggest that the optimality of the 90° temporal phase may be specific to the direction-discrimination task. Our experiments were based on a new procedure for measuring classical apparent motion thresholds (Hock, Kogan, and Espinoza, 1996, paper presented at ARVO). Two horizontally displaced dots are presented simultaneously against a darker background. The luminance ( L1) of one dot is always greater than that of the other ( L2), and the luminance values for the dots are exchanged on successive frames. Whether motion or stationarity is perceived depends on the background-relative luminance contrast (BRLC): ( L1- L2) divided by the difference between the average [( L1+ L2)/2] and background luminance. We found in the current study that motion thresholds depend on the temporal phase of the luminance variation at each location (rather than temporal asynchrony); the greater the phase difference (from 41° to 180°) the less the BRLC required for motion perception. At suprathreshold BRLC values, the perceived speed of apparent motion decreases with increased differences in temporal phase. The results are discussed in terms of Reichardt-type motion detection models.

scholarly journals Motion Direction Discrimination with Tactile Random-Dot Kinematograms

i-Perception ◽  
2021 ◽  
Vol 12 (2) ◽  
pp. 204166952110046
Author(s):  
Scinob Kuroki ◽  
Shin’ya Nishida
Keyword(s):  
Motion Detection ◽  
Visual Motion ◽  
Discrimination Performance ◽  
Sensory Function ◽  
Human Observer ◽  
Motion Sensing ◽  
Motion Direction ◽  
Low Level ◽  
Direction Discrimination ◽  
High Level

Motion detection is a fundamental sensory function for multiple modalities, including touch, but the mechanisms underlying tactile motion detection are not well understood. While previous findings supported the existence of high-level feature tracking, it remains unclear whether there also exist low-level motion sensing that directly detects a local spatio-temporal correlation in the skin-stimulation pattern. To elucidate this mechanism, we presented, on braille displays, tactile random-dot kinematograms, similar to those widely used in visual motion research, which enables us to independently manipulate feature trackability and various parameters of local motion. We found that a human observer is able to detect the direction of difficult-to-track tactile motions presented to the fingers and palms. In addition, the direction-discrimination performance was better when the stimuli were presented along the fingers than when presented across the fingers. These results indicate that low-level motion sensing, in addition to high-level tracking, contribute to tactile motion perception.

scholarly journals Spatio-temporal relays control layer identity of direction-selective neuron subtypes in Drosophila

2018 ◽  
Author(s):  
Holger Apitz ◽  
Iris Salecker
Keyword(s):  
Transcription Factor ◽  
Motion Detection ◽  
Visual Motion ◽  
Motion Direction ◽  
Core Domain ◽  
Lobula Plate ◽  
T Box ◽  
The Core ◽  
Spatio Temporal ◽  
Control Layer

Visual motion detection in sighted animals is essential to guide behavioral actions ensuring their survival. In Drosophila, motion direction is first detected by T4/T5 neurons. Their axons innervate one of four lobula plate layers. How T4/T5 neurons with layer-specific representation of motion-direction preferences are specified during development is unknown. We show that diffusible Wingless (Wg) between adjacent neuroepithelia induces its own expression to form secondary signaling centers. These activate Decapentaplegic (Dpp) signaling in adjacent lateral tertiary neuroepithelial domains dedicated to producing layer 3/4-specific T4/T5 neurons. T4/T5 neurons derived from the core domain devoid of Dpp signaling adopt the default layer 1/2 fate. Dpp signaling induces the expression of the T-box transcription factor Optomotor-blind (Omb), serving as a relay to postmitotic neurons. Omb-mediated repression of Dachshund transforms layer 1/2-into layer 3/4-specific neurons. Hence, spatio-temporal relay mechanisms, bridging the distances between neuroepithelial domains and their postmitotic progeny, implement T4/T5 neuron-subtype identity.

Developmental Constraints of Motion Detection Mechanisms in the Kitten

Perception ◽  
1977 ◽  
Vol 6 (5) ◽  
pp. 513-527 ◽  
Author(s):  
Jean-Marc Flandrin ◽  
Marc Jeannerod
Keyword(s):  
Superior Colliculus ◽  
Motion Detection ◽  
Visual Motion ◽  
Optokinetic Nystagmus ◽  
Ocular Dominance ◽  
Recovery Period ◽  
Direction Selectivity ◽  
Developmental Constraints ◽  
Superior Colliculus Neurons ◽  
Detection Mechanisms

The influence of deprivation procedures on the development of motion detection mechanisms has been studied in twenty-two kittens. Superior colliculus neurons did not acquire direction selectivity and normal ocular dominance in animals reared in the dark or in stroboscopic light. Neuron immaturity persisted in spite of a five week additional recovery period in normal conditions. Exposure to unidirectional visual motion for 10 h during the fifth week of postnatal age produced an asymmetric development of the two superior colliculi. Finally, unilateral neonatal ablation of visual cortex permanently impaired development of the ipsilateral superior colliculus. In the same or in different animals, development of optokinetic nystagmus, a typical visuomotor response, was similarly influenced by the global or selective deprivation procedures. These results suggest that motion detection mechanisms (both afferent and efferent) strongly depend upon constraints imposed by the visual world during the first weeks of life.

scholarly journals Electrophysiological aftereffects of high-frequency transcranial random noise stimulation (hf-tRNS): an EEG investigation

2021 ◽  
Author(s):  
Filippo Ghin ◽  
Louise O’Hare ◽  
Andrea Pavan
Keyword(s):  
High Frequency ◽  
Discrimination Task ◽  
Temporal Dynamics ◽  
Random Noise ◽  
Decomposition Analysis ◽  
Oscillatory Activity ◽  
Motion Direction ◽  
Time Frequency ◽  
Direction Discrimination ◽  
Transcranial Random Noise Stimulation

AbstractThere is evidence that high-frequency transcranial random noise stimulation (hf-tRNS) is effective in improving behavioural performance in several visual tasks. However, so far there has been limited research into the spatial and temporal characteristics of hf-tRNS-induced facilitatory effects. In the present study, electroencephalogram (EEG) was used to investigate the spatial and temporal dynamics of cortical activity modulated by offline hf-tRNS on performance on a motion direction discrimination task. We used EEG to measure the amplitude of motion-related VEPs over the parieto-occipital cortex, as well as oscillatory power spectral density (PSD) at rest. A time–frequency decomposition analysis was also performed to investigate the shift in event-related spectral perturbation (ERSP) in response to the motion stimuli between the pre- and post-stimulation period. The results showed that the accuracy of the motion direction discrimination task was not modulated by offline hf-tRNS. Although the motion task was able to elicit motion-dependent VEP components (P1, N2, and P2), none of them showed any significant change between pre- and post-stimulation. We also found a time-dependent increase of the PSD in alpha and beta bands regardless of the stimulation protocol. Finally, time–frequency analysis showed a modulation of ERSP power in the hf-tRNS condition for gamma activity when compared to pre-stimulation periods and Sham stimulation. Overall, these results show that offline hf-tRNS may induce moderate aftereffects in brain oscillatory activity.

Sign in / Sign up

Export Citation Format

Share Document