Purchase this article with an account.
Leonid Fedorov, Dominik Endres, Joris Vangeneugden, Martin Giese; Neurodynamical model for the multi-stable perception of biological motion.. Journal of Vision 2014;14(10):1007. doi: 10.1167/14.10.1007.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The perception of biological motion integrates information over time and likely is dependent on the fusion of multiple cues. Under normal conditions biological motion stimuli result in unambiguous percepts. Recent work, however, has shown that certain biological motion stimuli can result in multi-stable perception and spontaneous perceptual switching (Vanrie et al. 2004; Jackson et al., 2010; Vangeneugden et al. 2012). An example is that specific views of walker stimuli induce alternating percepts of locomotion direction. METHODS: We extended a physiologically-inspired dynamical neural model for the processing of body motion (Giese & Poggio, 2003) by inclusion of a joint dynamic representation of the temporal order of intermediate patterns and stimulus view. This representation is modeled by a two dimensional dynamic neural field. Inclusion of noise results in perceptual switching between two different travelling peak solutions that correspond to the alternating percepts. The model was trained with walker stimuli in different directions and tested with ambiguous views. RESULTS AND CONCLUSION: The model is able to reproduce spontaneous perceptual switching between different perceived directions of biological motion, and at least qualitatively accounts for the dependence of this bistability on view angle differences. More detailed simulations reproducing psychophysical data on the multistability are in progress. REFERENCES. Giese M. A., Poggio T. (2003). Nature Reviews Neuroscience, 4, 179-192. Vanrie J. et. al. (2004). Perception 33: 547-560. Vangeneugden J. et. el. (2012). Society for Neuroscience Meeting, 127.04. Jackson J., Blake R.(2010). Journal of Neuroscience, 30, 838-848.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only