- Location: School of Engineering, Featheringill Hall
- Room: 110
- Contact: Angel Gaither
- Email: email@example.com
- Phone: 615-322-0080
- Website: https://www.vanderbilt.edu/psychological_sciences/events/index.php
- Audience: Free and Open to the Public
Yanxi Liu, PhD
Professor of Computer Science & Engineering and Electrical Engineering
Penn State University
"Human Perception of Visual Patterns: A Nexus of Neuroscience, Psychology, Computer Science, & Mathematics"
Regularities with varying form and scale pervade our natural and man-made world. From insects to mammals, the ability to sense regular patterns has a neurobiological basis and has been observed in many levels of intelligence and behavior. Much of our understanding of the world is based on the perception, recognition and understanding of repeated patterns, generalized by the mathematical concept of symmetry and symmetry groups. Given the ubiquity of symmetry in both the physical and digital worlds, a computational model for symmetry-based regularity perception is especially pertinent for us to understand human visual pathways.
In this talk, I present our recent progress on understanding visual perception of regular patterns by humans, using visual stimuli organized by the 2D crystallographic groups and the measured neural responses in human participants (fMRI and EEG):
(1) We found that cortical area V3 has a parametric representation of the rotation symmetries in regular texture patterns, that is not present in either V1 or V2, the first discovery of a stimulus property that differentiates processing in V3 from that of lower-level areas. Parametric responses with respect to the degree of rotation symmetries were also seen in higher order ventral stream areas V4, VO1, and lateral occipital complex (LOC), but not in dorsal stream areas.
(2) Under a novel formalization of a class-topology concept, we apply machine learning algorithms on the EEG responses to the full 17 wallpaper groups and found a significant topological overlap between human perceived wallpaper patterns and the mathematically proven wallpaper subgroup hierarchy. In addition, behavior studies and deep learning models on the same visual stimuli support our findings.