Brain processing motion and curves

Motion and Perception: How Your Brain Interprets Movement

Nico Varela monogram
Nico Varela in Mind & Education October 2026 • 4 min read.

"Uncover the fascinating link between curvature coding and motion direction to understand how your brain processes visual motion."


Our perception of the world relies heavily on two key visual processes: shape recognition and motion detection. For a long time, scientists believed these were handled by separate pathways in the brain—shape primarily through the ventral stream and motion via the dorsal stream. However, mounting evidence suggests that these processes are far more interconnected than previously thought.

Neurophysiological studies, for instance, have revealed that neurons in areas V4 and IT (both ventral stream regions) respond to shapes defined solely by the coherent motion of random dots. Additionally, anatomical studies have demonstrated strong connections between the motion-sensitive dorsal area MT and the contour/shape-sensitive ventral area V4, painting a picture of constant communication and integration.

This article delves into a specific aspect of this interaction: how curvature, a fundamental element of shape, is encoded in relation to motion direction. By examining the motion direction tuning of curvature processing, we can better understand the mechanisms that allow us to perceive and interact with a dynamic environment.

AI Search Multiple angles on this topic

Current Statistics & Impact

The human visual system processes motion through a dual-channel architecture that detects both first-order and second-order motion signals. The eye moves constantly, fixating new scene locations every 300 milliseconds, creating a continuous stream of changing visual input that makes motion perception a core component of visual processing. Machine learning models trained on moving images can recapitulate key properties of biological motion direction and speed processing, offering computational insights into motion misperception. Understanding sampled motion effects on perception is increasingly important for designing future display technologies.

Standard Approach & Limitations

The standard model of vision faces challenges explaining puzzling data in visual neuroscience and perception experiments. Visual motion signals drive both motion perception and oculomotor responses, yet a fundamental question remains whether perceptual and oculomotor processing share common mechanisms or operate independently across the visual motion hierarchy. Motion perception involves inferring speed and direction from visual, vestibular, and proprioceptive inputs—a computationally difficult problem that has proven challenging to explain in terms of neural processing. New models like INRF have been developed to overcome limitations of traditional motion sensor models.

Historical Perspective

Research into luminance changes affecting motion perception has yielded extensive information about first-order motion processes in the retina and second-order processes in the visual cortex. Motion perception involves a substantial network of brain regions contributing to functions including object motion perception, self-motion, guidance of eye and hand movements, and perception of moving two- and three-dimensional object forms. A fundamental open question concerns the evolutionary origins of specialized motion perception forms such as second-order motion, structure-from-motion, and biological motion—assuming dedicated mechanisms exist, which remains unsettled. Visual motion signals serve dual roles in driving perception and eliciting oculomotor responses.

What is Curvature Coding?

Brain processing motion and curves

Curvature coding refers to the way our brains process and interpret curves and contours within visual stimuli. It's a critical aspect of shape recognition, allowing us to differentiate between objects and understand their forms. Researchers have long been intrigued by how the brain achieves this, particularly in the context of movement.

To explore the link between curvature coding and motion direction, scientists often use shape after-effects, such as the shape-frequency after-effect (SFAE) and the shape-amplitude after-effect (SAAE). These after-effects occur when adapting to a specific shape causes a shift in the perceived frequency or amplitude of a subsequently viewed test shape. By manipulating the motion of these shapes, researchers can gain insights into the motion sensitivity of the underlying curvature-coding mechanisms.

Here are the key questions that researchers aim to address:
  • Are curvature-encoding mechanisms selective for global motion direction (the overall direction of a shape)?
  • Are these mechanisms sensitive to shape temporal frequency (the rate at which the shape changes over time)?
  • How do luminance properties, such as spatial frequency (the fineness of detail) and temporal frequency (the rate at which luminance changes), affect the interaction between motion and curvature?
  • Do local motion cues (motion of individual components within a shape) play a role in curvature coding?
AI Search Multiple angles on this topic

Latest Research and Reviews

Recent advances in computational modeling and neural network approaches are providing new insights into how the brain processes visual motion. Machine learning models trained on naturalistic motion stimuli are beginning to replicate key properties of biological motion perception, including direction and speed tuning. These computational approaches complement traditional psychophysical and neurophysiological methods, offering testable predictions about neural mechanisms underlying motion processing.

Counter Arguments and Failures

Visual motion processing faces fundamental challenges when incoming information is interrupted by occlusion and eyeblinks, creating perceptual gaps that the brain must resolve through extrapolation and prediction mechanisms. The predominant view that motion and form processing operate through separate, independent modules has been challenged by evidence of interactions between these pathways. Motion processing traditionally involves direction-selective filtering followed by integration to solve the aperture problem, while form processing uses orientation- and size-selective receptive fields, but the independence of these streams is increasingly questioned.

Comparative Analysis

Research comparing perceptual and pursuit eye movement responses to visual motion reveals both similarities and dissociations, suggesting common and separate principles guide perceptual and motor responses. Computational models from computer vision, particularly optical-flow estimation models, are being tested against human psychophysical phenomena to gain insights into the computational mechanisms of biological visual motion processing. These comparative approaches help identify which aspects of motion perception can be explained by general-purpose algorithms versus specialized biological mechanisms.

To answer these questions, a series of experiments were conducted using moving sine-wave-shaped contours. These contours were designed to drift within a fixed stimulus window, either with global motion (the entire contour moving) or local motion (only the internal elements moving). The researchers then measured the SFAEs and SAAEs under various conditions, manipulating motion direction, shape temporal frequency, and luminance properties.

Key Findings and Implications

The experiments revealed several key findings: Global Motion Selectivity: Both SFAEs and SAAEs exhibited selectivity for global motion direction, meaning the after-effects were stronger when adaptor and test contours moved in the same direction. Temporal Frequency Dependence: For the same-motion-direction condition, both after-effects increased with shape temporal frequency. Luminance Influence: Luminance temporal frequency partially explained the increase in after-effect magnitude with shape temporal frequency. Lack of Local Motion Selectivity: Local motion after-effects did not show selectivity for motion direction. These results suggest that curvature is encoded by mechanisms that are selective for motion direction, and that this directional selectivity is most apparent when the stimulus provides sufficient spatio-temporal coverage. In simpler terms, when your brain has enough visual information over a short period, it can better determine the direction in which a curve is moving.

AI Search Multiple angles on this topic

Synthesis & Expert Commentary

The field of motion perception research is converging on a more integrated understanding that bridges computational, neural, and behavioral levels of analysis. Evidence from multiple methodologies suggests that motion processing is not a single monolithic system but involves interacting pathways with both shared and specialized components. This synthesis highlights the importance of considering motion perception as an active, predictive process rather than a passive feedforward computation.

Future Outlook & Next Frontiers

Future research will likely focus on how motion perception integrates with other cognitive functions such as attention, memory, and decision-making in naturalistic environments. Advances in neuroimaging, computational modeling, and virtual reality technologies promise to reveal how the brain handles motion processing during active behavior. Understanding individual differences in motion perception and their links to developmental and neurological conditions represents another key frontier.

Broader Context & Systemic Challenges

Motion perception plays a fundamental role in daily visual experience, critical for computing the speed and direction of moving objects and controlling body and eye movements. Visual neuropsychology reveals that vision is not a passive function but dynamically interacts with other domains of human life, impacting typical and atypical development at systemic and integrated levels. These broader relationships highlight how visual competences extend beyond basic perception to influence complex cognitive and motor functions.

The Human Element & Real-World Impact

Visual perceptual learning can enhance internal processing efficiency for non-directional motion perception, with training effects dependent on stimulus characteristics. These findings demonstrate that motion perception is not fixed but can be improved through experience, with potential applications for rehabilitation and performance enhancement. The transfer mechanisms of perceptual learning suggest that targeted training could benefit real-world tasks requiring motion discrimination.

What trajectory straightening can-and cannot-tell us

The experiments behind the temporal straightening hypothesis only partly reproduce the conditions that shape neural trajectories during natural viewing. The measured neural and perceptual effects also differed in size, so the neural result should not be treated as a direct match for perception.[1]

One proposed explanation is that the most responsive V1 neurons may account for the perceptual effect, but the experiment may not have recorded them because its stimuli were not tailored to those neural populations. This is a possible explanation, not a demonstrated cause of the gap.[1]

How motion and form evidence fits together

For static and dynamic Glass patterns, the cited account distinguishes detecting the orientation of individual dipoles from pooling those local cues to extract an overall shape. This offers a way to frame form perception as a progression from local signals to a global pattern.[2]

Biological-motion perception may involve combining points into a body form, but a single informative trajectory or partial representation may sometimes be enough. This makes the role of whole-body integration dependent on what motion cues are present.[3]

Motion direction sensitivity can vary by visual-field location: the reported group-level difference in sensitivity was not statistically significant, while an interaction reflected opposing effects at and away from the primary meridians. The overall comparison therefore does not support a simple uniform difference across locations.[4]

What further evidence would need to show

  • How closely would neural trajectory effects match perceptual effects if experiments captured the conditions of natural viewing more fully? The current test only partly engages those conditions, and the reported neural and perceptual effects differ in magnitude.[1]
  • How are local motion signals parsed into larger patterns, and how do computational descriptions of that hierarchy relate to neural processing?[5]

Questions about measuring motion perception

Can motion detection be measured consistently while a person is walking?

Repeated measurements showed strong correlations in both walking and standing conditions, with good agreement between repeated measures. This suggests the measure can be reliable across those conditions.[6]

Terms for describing form and motion

Bayesian framework
A way to combine information while accounting for its reliability[7]

Sources

  1. Primary visual cortex straightens natural video trajectories (pmc.ncbi.nlm.nih.gov)
  2. Investigating the Interaction Between Form and Motion Processing: A Review of Basic Research and Clinical Evidence (pmc.ncbi.nlm.nih.gov)
  3. Perception of biological motion in visual agnosia (pmc.ncbi.nlm.nih.gov)
  4. Asymmetries in the discrimination of motion direction around the visual field (pmc.ncbi.nlm.nih.gov)
  5. Hierarchical Vector Analysis of Visual Motion Perception. (doi.org)
  6. Visual motion detection thresholds can be reliably measured during walking and standing (pmc.ncbi.nlm.nih.gov)
  7. A comparison of form processing involved in the perception of biological and nonbiological movements (pmc.ncbi.nlm.nih.gov)

About this Article -

Written with AI assistance from published research, and reviewed by the Mystum team. See our About page for more information.

Everything You Need To Know

1

What is curvature coding, and why is it important for understanding how we see movement?

Curvature coding is the brain's method for processing and interpreting the curves and contours within visual stimuli. This is a fundamental aspect of shape recognition, enabling us to distinguish between objects and grasp their forms. It's crucial for motion perception because the brain utilizes curvature information to understand how shapes move and change over time. Scientists investigate this using shape after-effects like shape-frequency after-effect (SFAE) and shape-amplitude after-effect (SAAE), which help uncover the mechanisms of motion direction tuning of curvature processing.

2

How do the brain's shape recognition and motion detection processes interact, and where do key brain regions like V4, IT, and MT come into play?

The article highlights that shape recognition (primarily handled by the ventral stream) and motion detection (primarily handled by the dorsal stream) are interconnected. Brain areas V4 and IT (both in the ventral stream) respond to shapes defined by motion. Anatomical studies show strong connections between the motion-sensitive dorsal area MT and the contour/shape-sensitive ventral area V4. These connections indicate constant communication and integration between shape and motion processing. Curvature coding is one area where these two processes interact.

3

Can you explain the experiments mentioned that explore the link between curvature coding and motion direction, and what key questions were they designed to answer?

The experiments involved using moving sine-wave-shaped contours with global or local motion. Researchers measured the SFAEs and SAAEs under various conditions. The key questions addressed were: Are curvature-encoding mechanisms selective for global motion direction? Are these mechanisms sensitive to shape temporal frequency? How do luminance properties affect motion and curvature interaction? Do local motion cues play a role? These experiments aimed to understand how our brains perceive and interpret moving shapes, focusing on curvature's role.

4

What were the main findings of the experiments, and what do they tell us about how the brain processes movement in relation to shape?

The experiments revealed that both SFAEs and SAAEs showed selectivity for global motion direction, meaning the after-effects were stronger when adaptor and test contours moved in the same direction. They also showed that for the same-motion-direction condition, the after-effects increased with shape temporal frequency. Luminance temporal frequency partially explained this increase. Local motion after-effects didn't show selectivity for motion direction. These results suggest that curvature is encoded by mechanisms selective for motion direction and that directional selectivity is most apparent with sufficient spatio-temporal coverage.

5

Why is understanding curvature coding and its relationship to motion direction important for our understanding of visual perception, and what are the broader implications?

Understanding curvature coding is crucial because it reveals how our brains extract meaningful information from the visual world. It helps explain how we perceive and interact with a dynamic environment. It provides insights into how the brain integrates shape and motion, demonstrating that these processes are not entirely separate. The broader implications extend to fields like computer vision and robotics, where understanding how the brain processes visual information can inspire more sophisticated algorithms and systems for object recognition and motion tracking. The knowledge helps us comprehend the underlying mechanisms of visual perception and potential impairments in individuals with visual processing disorders.

Newsletter Subscribe

Subscribe to get the latest articles and insights directly in your inbox.