Sorry, your browser doesn't support canvas element !
Sorry, your browser doesn't support canvas element !
Shape perception from illusory contours
4 pacmen forming a fat or thin illusory shape
50
% of square side
5
deg
100
%
100
%
0
Hz
2
pixels
50
%
v1.0
© 2020 KyberVision - Innovation in Vision Sciences
Send Feedback
Kanizsa-like shapes have been widely used to investigate the spatial and temporal properties of illusory contours and amodal completion in shape discrimination tasks, and unveil the mechanisms that recover object size, shape, and position from sparsely segregated edge elements.

Kellman & Shipley (1991) described a theory explaining the perception of partly occluded objects and illusory figures, from both static and kinematic information, in a unified framework. However it remained unclear how the cortex implements the visual processing underlying the perception of illusory contours and amodal completion.

Neural responses to illusory contour were found in earlier cortical areas of macaque monkeys like V1, in midlevel visual areas like V4 as well as in higher cortical areas like the inferior temporal cortex (IT) suggesting that the computational power to detect illusory contours and distinguish figure from ground is available early in the visual cortex. However the temporal sequence of the events suggests that this computation involves inter-cortical interaction, and that early perceptual organization is likely to be an interactive process.

Studies using brain imaging techniques like fMRI, MEG, EEG or TMS, have demonstrated since that multiple cortical regions are also activated in the human visual cortex when static or moving Kanizsa-like illusory contours are presented. They showed that both earlier visual areas like V1/V2 and higher-level visual area like LO are critically involved in perceptual completion but with different timing. These results suggest that responses in earlier visual areas may be influenced by top-down modulation from higher areas, which contribute to perceptual completion through neural feedback to V1/V2 areas.

These results may have some clinical implications as schizophrenia patients exhibit abnormal electrophysiological signatures during Kanizsa shape perception tasks. While it remains unclear how these abnormalities are manifested behaviorally and whether they arise from early or late levels in visual processing, a recent study suggests that it is a later conceptually-mediated shape integration stage which is compromised in schizophrenia.

References:

  Kanizsa (1976) Subjective Contours. Scientific American 234(4):48–52

  von der Heydt et al. (1984) Illusory contours and cortical neuron responses. Science 224: 1260–1262

  Kellman & Shipley (1991) A theory of visual interpolation in object perception. Cogn Psychol 23:141

  Peterhans & von der Heydt (1993) Subjective contours - bridging the gap between psychophysics and physiology . Trends in Neurosciences 14(3):112–119

  Grosof et al. (1993) Macaque V1 neurons can signal “illusory” contours. Nature 365:550–552

  Ringach & Shapley (1996) Spatial and temporal properties of illusory contours and amodal boundary completion. Vision Research 36:3037–3050

  Mendola et al. (1999) The Representation of Illusory and Real Contours in Human Cortical Visual Areas Revealed by Functional Magnetic Resonance Imaging. Journal of Neuroscience 19(19):8560–8572

  Gold et al. (2000) Deriving behavioural receptive fields for visually completed contours. Current Biology 10:663–666

  Seghier et al. (2000) Moving illusory contours activate primary visual cortex: an fMRI study. Cereb Cortex 10:663–670

  Lee & Nguyen (2001) Dynamics of subjective contour formation in the early visual cortex. PNAS 98:1907–1911

  Halgren et al. (2003) Cortical activation to illusory shapes as measured with magnetoencephalography. Neuroimage 18:1001–1009

  Stanley & Rubin (2003) fMRI activation in response to illusory contours and salient regions in the human lateral occipital complex. Neuron 37:323–331

  Murray et al. (2006) Boundary Completion Is Automatic and Dissociable from Shape Discrimination. Journal of Neuroscience 26:12043–12054

  Sáry et al. (2008) The representation of Kanizsa illusory contours in the monkey inferior temporal cortex. European Journal of Neuroscience 28:2137–2146

  Shpaner et al. (2009) Early processing in the human lateral occipital complex is highly responsive to illusory contours but not to salient regions. European Journal of Neuroscience 30:2018–2028

  McMains & Kastner (2010) Defining the units of competition: influences of perceptual organization on competitive interactions in human visual cortex. J Cogn Neurosci 22:2417– 2426

  Wokke et al. (2013) Confuse your illusion: feedback to early visual cortex contributes to perceptual completion. Psychological Science 24:63–71

  Cox et al. (2013) Receptive field focus of visual area V4 neurons determines responses to illusory surfaces. PNAS 110:17095–17100

  Wyatte et al. (2014) Early recurrent feedback facilitates visual object recognition under challenging conditions. Front Psychol 5:760–10

  Keane et al. (2014) Late, not early, stages of Kanizsa shape perception are compromised in schizophrenia. Neuropsychologia 56:302–311

  Kok & de Lange (2014) Shape Perception Simultaneously Up- and Downregulates Neural Activity in the Primary Visual Cortex. Current Biology 24:1531–1535

  Keane et al. (2020) Network mechanisms of shape completion. bioRxiv Preprint
Here is the math behind this stimulus:

  ecc = side/2
  rad2 = (0.01*radius*ecc)^2
  angle = (pi/10)*cos(2*pi*time)
  xm = x-ecc; xp = x+ecc; ym = y-ecc; yp = y+ecc
  disk1 = (xm^2+ym^2)<rad2 & mod(abs(atan(ym,xm)+3*pi-angle),2*pi)>pi/2
  disk2 = (xp^2+ym^2)<rad2 & abs(atan(ym,xp)+pi/4+angle)>pi/4
  disk3 = (xm^2+yp^2)<rad2 & mod(abs(atan(yp,xm)+3*pi/2+angle),2*pi)>pi/2
  disk4 = (xp^2+yp^2)<rad2 & abs(atan(yp,xp)-pi/4-angle)>pi/4
  disks = 0.01*alpha*(disk1+disk2+disk3+disk4)
  back = 0.5*(1+0.01*cnt*unoise(x,round(noisetf*time),noisegran))
  z = back*(1 - disks) + disks*lum/100

Where:

  'radius' is the pacmen radius (in % of square side)
  'side' is the square side (in deg)
  'lum' is the pacmen luminance (in %)
  'alpha' is the opacity (in %)
  'noisetf' is the noise temporal frequency (in Hz)
  'noisegran' is the noise granularity (in pixels)
  'cnt' is the noise contrast (in %)
The whole stimulus is generated in real-time using a GLSL shader that runs right inside your WebGL-compatible browser. The plain Math behind the stimulus was converted to this optimized GLSL shader using the new Psykinematix Pro Edition. Translation to Matlab and Python code is also possible !

This whole widget was also fully generated using Psykinematix Pro Edition. The parameters that control the stimulus properties through the sliders are the same as the ones you would define as dependent or independent variables when using the stimulus in an actual psychophysical experiment run in Psykinematix. The widget creation is otherwise fully customizable with your own logo, copyright, links, etc.

To learn more about the widget creation, click on the above "Made With" button !
v1.0
© 2020 KyberVision - Innovation in Vision Sciences