Imaging Science Seminar: New Approach to Stereo Vision

Center for Imaging Science Seminar
New Approach to Stereo Vision

Dr. Paul Linton
Centre for Applied Vision Research, University of London

Register Here for Zoom Link

Abstract:
Virtual and augmented reality promise to transform the central question of display technology from “image distortion” to “spatial distortion”, asking: “How do distortions in displays affect not just the image we see, but the very 3D space we experience?” To answer this question, we need to have a good model of stereo vision (depth perception based on the different views of the two eyes). However, the argument of my talk is that stereo vision is poorly understood. Traditional models of stereo vision for displays (Woods et al., 1993) treat stereo vision as trying to triangulate the position of points in the world. So quantifying stereo vision distortions relies on triangulating the “misperceived” position of points in the world given their misprojection in the two eyes (Woods et al., 1993). But there is little evidence that the human visual system works in this way, and we readily experience depth in stereoscopic displays that violate this account. So, in this talk, I develop an alternative to the traditional triangulation model.

Bio:
Paul Linton is a neuroscientist and philosopher specializing in 3D vision. He received his PhD in 2021 from the Center for Applied Vision Research, City, University of London, where his research challenged our understanding of distance perception by showing the visual system is unable to triangulate distance using the two eyes. He was also part of the DeepFocus team at Meta Reality Labs. Paul is the author of The Perception and Cognition of Visual Space (Palgrave Macmillan, 2017). Before vision science, he was a stipendiary lecturer in law at St Hilda’s College, Oxford University, and a teaching fellow in philosophy at University College London. As a Presidential Scholar, Paul will develop his new two-stage theory of 3D vision using the latest techniques in machine learning and fMRI in the hope of explaining how we experience the 3D world.

Intended Audience:
All are welcome.

To request an interpreter, please visit myaccess.rit.edu