Search Papers | Poster Sessions | All Posters
Poster B133 in Poster Session B - Thursday, August 8, 2024, 1:30 – 3:30 pm, Johnson Ice Rink
The Representational Organization of Static and Dynamic Visual Features in the Human Cortex
Hamed Karimi1 (), Jeff Wang1, Stefano Anzellotti1; 1Boston College
Visual information consists of static and dynamic properties. How is their representation organized in the visual system? Static information has been associated with ventral temporal regions and dynamic information with lateral and dorsal regions. However, investigating the representation of static and dynamic information is complicated by the correlation between static and dynamic features of continuous visual input. Recent work addressed this challenge by using point-light displays and kinematograms to isolate motion information, but such stimuli might not capture the rich dynamics contained in realistic videos. Here, we separated static and dynamic features in realistic videos using two-stream deep convolutional neural networks and used them in conjunction with fMRI and representational similarity analysis to investigate the representation of static and dynamic information in the visual system. First, consistent with recent findings, we showed that both static and dynamic features are encoded across all visual streams. Second, we found that brain streams represent shared as well as unique static and dynamic visual information.
Keywords: action recognition visual pathways static and dynamic visual features deep neural networks