Abstract:Successful recognition of objects in cluttered scenes entails integration of fragmented images across space and time. Previous work focused on the integration of visual information across space. However, little is known about spatiotemporal integration that involves storing discrete image components in short-term memory and binding them across space and time. Our psychophysical work showed that collinear contours are reliably detected and identified when moving behind a narrow slit, suggesting that the spatiotemporal contour integration may not depend on V1 horizontal connections that are thought to underlie spatial contour integration. In the current study we used fMRI to investigate the neural mechanisms that mediate spatiotemporal contour integration under slit-viewing. The stimuli comprised arrays of randomly orientated Gabor patches, either with (contour stimuli) or without (random stimuli) embedded contour paths. The contour stimuli contained five embedded contours each comprising 4~9 collinear Gabor elements. The global orientation of the contours was jittered between 30°-45° or 135°-150°. In a blocked design the contour or random stimuli moved behind a slit whose width (0.75°) was equal to the average inter-element distance. As a result, at any given moment only one or parts of two contour elements could be seen through the slit. Within a block of trials a small proportion (12.5%) of stimuli contained vertical contours as targets. Observers were instructed to detect the target stimuli. A GLM analysis showed stronger fMRI responses to contour stimuli than random stimuli in higher dorsal visual areas (V7, V3B/KO, VIPS) and in the LOC, suggesting that spatiotemporal integration involves interactions between ventral and dorsal visual areas. We hypothesize that the dorsal areas may contribute to binding moving contour elements across space and time, and that the ventral regions may generate the integrated percept of the global contour.

Contour Detection Model with Multi-Scale Integration Based on Non-Classical Receptive Field

Contour Detection Model Based On The Combination Of Surround Facilitation And Inhibition

Multi-scale Image Analysis Based on Non-Classical Receptive Field Mechanism.

Contour detection based on the interactive response and fusion model of bilateral attention pathways

Image Contour Detection Based on Visual Pathway Information Transfer Mechanism

A contour detection method based on the projective coding model of the visual cortex information flow

Application of binocular disparity and receptive field dynamics: A biologically-inspired model for contour detection

A Neural Model of Visual Contour Integration.

The Neural Mechanism of Visual Contour Integration

An Integration Model Based on Non-classical Receptive Fields.

Vis-CRF, A Classical Receptive Field Model for VISION

Visually-salient contour detection using a V1 neural model with horizontal connections

A Neural Model of Contour Integration in the Primary Visual Cortex

A multiple sub-regions design of non-classical receptive field

Scale-invariant contour segment context in object detection.

A feature aggregation network for contour detection inspired by complex cells properties

Adaptive Contour Model for Real-Time Foreground Detection

A convex neighbor-constrained active contour model for image segmentation

Contour Detection Based on Self-Organizing Feature Clustering

Neighbor-constrained Active Contours Without Edges.

Spatiotemporal contour integration in the human brain