Host: Prof. Shuguang Kuai, East China Normal University
Abstract:
The stereokinetic phenomenon is one of the most striking demonstrations of the visual system’s ability to construct three-dimensional structure from highly ambiguous motion signals. Although known for more than a century and famously explored by artists such as Marcel Duchamp, the computational principles underlying stereokinetic perception remain poorly understood.
In this talk, I will discuss a series of studies examining how the visual system infers shape from rotating two-dimensional patterns that contain little or no explicit depth information. I will first place stereokinesis within the broader context of shape-from-motion perception and perceptual regularization principles. I will then present quantitative work showing that perceived stereokinetic geometry can be predicted from minimal-deformation constraints, followed by more recent results characterizing how the inferred depth percept scales with stimulus geometry and varies across observers.
Together, these findings suggest that stereokinetic perception provides an unusually transparent window into the lawful geometric constraints the visual system imposes on ambiguous motion stimuli.
Biography:
Zili Liu studied physics at Peking University (1981 - 85). He received his MS in Applied Math, and PhD in Cognitive Science, both from Brown University, USA. He is a professor at UCLA's Cognitive Psychology and Brain Research Institute.
This event is open to the NYU Shanghai, East China Normal University, and Neuroscience community.