Abstract
Classic motion energy models are able to predict a wide range of physiological and behavioral aspects of motion perception in humans. Whether these models can be used as a basis for higher-level tasks, such as moving object segmentation, has however hardly been explored yet. Here, we present a model that combines a motion energy representation with recent computer vision approaches for figure-ground segmentation of naturalistic stimuli. We find that unlike established motion segmentation models but similar to humans, our model generalizes to random-dot stimuli when only trained on RGB videos.
Keywords
motion segmentation, motion energy, figure-ground segregation
Start Date
17-5-2024 9:00 AM
End Date
17-5-2024 10:00 AM
Recommended Citation
Tangemann, Matthias; Kümmerer, Matthias; and Bethge, Matthias, "Fusing Classic Motion Energy Models and Deep Learning for Coarse-to-fine Moving Object Segmentation" (2024). MODVIS Workshop. 1.
https://docs.lib.purdue.edu/modvis/2024/mv/1
Fusing Classic Motion Energy Models and Deep Learning for Coarse-to-fine Moving Object Segmentation
Classic motion energy models are able to predict a wide range of physiological and behavioral aspects of motion perception in humans. Whether these models can be used as a basis for higher-level tasks, such as moving object segmentation, has however hardly been explored yet. Here, we present a model that combines a motion energy representation with recent computer vision approaches for figure-ground segmentation of naturalistic stimuli. We find that unlike established motion segmentation models but similar to humans, our model generalizes to random-dot stimuli when only trained on RGB videos.