Cue combination in the motion correspondence problem

Abstract

Image motion is a primary source of visual information about the world. However, before this information can be used the visual system must determine the spatio-temporal displacements of the features in the dynamic retinal image, which originate from objects moving in space. This is known as the motion correspondence problem. We investigated whether cross-cue matching constraints contribute to the solution of this problem, which would be consistent with physiological reports that many directionally selective cells in the visual cortex also respond to additional visual cues. We measured the maximum displacement limit (D max ) for two-frame apparent motion sequences. D max increases as the number of elements in such sequences decreases. However, in our displays the total number of elements was kept constant while the number of a subset of elements, de¢ned by a di¡erence in contrast polarity, binocular disparity or colour, was varied. D max increased as the number of elements distinguished by a particular cue was decreased. D max was a¡ected by contrast polarity for all observers, but only some observers were in£uenced by binocular disparity and others by colour information. These results demonstrate that the human visual system exploits local, cross-cue matching constraints in the solution of the motion correspondence problem.

Mechanisms for similarity matching in disparity measurement

“…Correspondence matching biases have been found based on similarity of contrast ( Anderson and Nakayama, 1994 ; Smallman and McKee, 1995 ; Goutcher and Mamassian, 2005 ), contrast polarity ( Watanabe, 2009 ), luminance ( Goutcher and Hibbard, 2010 ), color ( den Ouden et al, 2005 ), orientation, motion direction, and speed ( van Ee and Anderson, 2001 ). Similar matching constraints have also been demonstrated in motion perception, for which an analogous matching problem exists ( Hibbard et al, 2000 ). Such results provide compelling evidence to support computational assertions of the importance of feature similarity.…”

Influence of Correspondence Noise and Spatial Scaling on the Upper Limit for Spatial Displacement in Fully-Coherent Random-Dot Kinematogram Stimuli

“…Each field of dots was presented for 157 ms (11 video frames). No frames separated the two fields (interstimulus interval (ISI) was 0 ms), as performance for identifying motion direction has been shown to be best at this ISI, regardless of dot-density, when the dot-displacement is smaller than 50 arcmin ( [42] , see also [36] ). Each dot in the two fields was a square of 4×4 arcmin, with a luminance of 76.8 cd/m 2 against a background luminance of 0.1 cd/m 2 .…”

“…In order to successfully identify the direction of motion from adjacent frames of an RDK, the visual system must solve the correspondence problem , i.e. it must determine the dot-displacement between frames that maximises the overlap between dots in adjacent frames ( [1] , [30] – [36] , see also [37] – [39] ). The correspondence problem exists because any dot in one frame can be paired with any dot in the adjacent frame, but only a small proportion of the potential pairs are signal pairs.…”

“…The issue of correspondence noise in RDK stimuli has also been studied with the Dmax paradigm (e.g. [36] , [40] , [41] ). The purpose of this paper is to further investigate the role of correspondence noise in motion perception for RDK stimuli, particularly with regard to its influence on Dmax.…”

“…of the dots between frames (e.g. [36] , [44] ). The general finding of these studies is that manipulations that reduce the probability of dot-mismatch across frames result in increases in Dmax.…”