Scientific Article
previous document  unige:17394  next document
add to browser collection
Title

Probabilistic Motion Estimation Based on Temporal Coherence

Authors
Yuille, Alan L.
Grzywacz, Norberto M.
Published in Neural Computation. 2000, vol. 12, no. 8, p. 1839-1867
Abstract We develop a theory for the temporal integration of visual motion motivated by psychophysical experiments. The theory proposes that input data are temporally grouped and used to predict and estimate the motion flows in the image sequence. This temporal grouping can be considered a generalization of the data association techniques that engineers use to study motion sequences. Our temporal grouping theory is epxressed in terms of the Bayesian generalization of standard Kalman flitering. To implement the theory, we derive a parallel network that shares some properties of cortical networks. Computer simulations of this network demonstrate that our theory qualitatively accounts for psychophysical experiments on motion occlusion and motion outliers. In deriving our theory, we assumed spatial factorizability of the probability distributions and made the approximation of updating the marginal distributions of velocity at each point. This allowed us to perform local computations and simplified our implementation. We argue that these approximations are suitable for the stimuli we are considering (for which spatial coherence effects aer negligible).
Keywords Kalman filterBayesian formulationMotion flow fieldKolmogorov forward equationFokker-Plank equationTemporal groupingTemporal coherencePsychophysical motion phenomenaVelocity estimationMotion occlusionMotion outlier
Identifiers
Full text
Citation
(ISO format)
BURGI, Pierre-Yves, YUILLE, Alan L., GRZYWACZ, Norberto M. Probabilistic Motion Estimation Based on Temporal Coherence. In: Neural Computation, 2000, vol. 12, n° 8, p. 1839-1867. https://archive-ouverte.unige.ch/unige:17394

213 hits

116 downloads

Update

Deposited on : 2011-11-14

Export document
Format :
Citation style :