Perrone, J.A. & Krauzlis, R.J. (2012). A contrast-sensitive, redundancy reduction mechanism acting on MT neurons can explain global motion direction biases without the need for Bayesian priors. Journal of Vision, 12(9), 932.
Permanent Research Commons link: http://hdl.handle.net/10289/7350
Introduction: The perceived global direction of moving objects can be influenced by the contrast of the object (Weiss, Simoncelli & Adelson, Nature Neuroscience, 2002). There is currently no detailed, neural-based explanation for how this could occur. Weiss et al., proposed an ideal Bayesian observer model that included a physiologically unspecified ‘low speed’ prior. We have recently developed a new velocity code for extracting image velocity from small groups of MT neurons (Perrone & Krauzlis, VSS 2011). The code includes a stage of local spatial inhibition between MT neurons designed to reduce the amount of redundant signals passed onto the global motion integration stage (MST). The inhibition is made dependent upon the contrast of the stimulus by exploiting the fact that some MT neurons change their speed tuning (V shifts to ∼ .5V) at low contrast (Krekelberg et al., J. Neurosci., 2006). An inhibitory signal based on the difference in output from two such MT units tuned to V and 2V will be high at high contrast, but will switch off at low contrast when the input speed is V. Methods: We tested our velocity code using high- and low-contrast rhombus stimuli that have been shown to produce contrast dependent biases in global motion estimates (Weiss, et al., 2002). Results: The model produced contrast dependent estimates of global motion direction that match the perceptual results. Low contrast rhomboids tended to produce estimates orthogonal to the major axis of the rhombus whereas high contrast rhomboids produced estimates closer to the veridical direction. Our simulations indicate that at low contrast the spatial inhibitory mechanisms between MT neurons may be switching off causing an increase in the density of responses signalling motion orthogonal to the edges. Conclusion: Our model provides a biologically plausible mechanism by which the Weiss et al., Bayesian prior could be implemented.
Association for Research in Vision & Ophthalmology