English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Poster

Neural model for the learning of biological motion

MPS-Authors
/persons/resource/persons84832

Jastorff,  J
Max Planck Institute for Biological Cybernetics, Max Planck Society;
Department Human Perception, Cognition and Action, Max Planck Institute for Biological Cybernetics, Max Planck Society;

/persons/resource/persons84787

Giese,  MA
Max Planck Institute for Biological Cybernetics, Max Planck Society;
Department Human Perception, Cognition and Action, Max Planck Institute for Biological Cybernetics, Max Planck Society;

Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Jastorff, J., & Giese, M. (2002). Neural model for the learning of biological motion. Poster presented at 25th European Conference on Visual Perception (ECVP 2002), Glasgow, UK.


Cite as: http://hdl.handle.net/11858/00-001M-0000-0013-DF60-7
Abstract
Several experimental results suggest that biological movements are encoded on the basis of learned template patterns. Recent theoretical work has shown that a biologically plausible hierarchical neural model that encodes biological movements in terms of learned prototypical patterns provides a consistent explanation for many experimental results on biological motion perception [Giese and Poggio, 2002 (submitted)]. The model postulates that biological-motion patterns are encoded in terms of sequences of complex shapes and optic-flow-field patterns in the ventral and dorsal pathway. The underlying neural circuits predict future form and optic-flow patterns from the previous stimulus sequence. Electrophysiological experiments (eg Markram et al, 1997 Science 275 213 - 215) have provided evidence that Hebbian plasticity in the cortex depends critically on the relative timing between the presynaptic and postsynaptic spikes. We tested how far such time-dependent Hebbian plasticity is an appropriate mechanism for the learning of sequence selectivity in the recognition of biological movements. In addition, we tried to derive psychophysically testable predictions that would allow us to validate the predictive coding hypothesis that underlies our model.