Probabilistic Motion Estimation Based on Temporal Coherence
This work addresses motion perception in vision science, offering a theoretical framework that generalizes data association techniques, but it appears incremental as it builds on existing Kalman filtering methods.
The authors tackled the problem of visual motion estimation by developing a Bayesian theory based on temporal coherence, which accounts for motion occlusion and outliers in psychophysical experiments through computer simulations.
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 used by engineers to study motion sequences. Our temporal-grouping theory is expressed in terms of the Bayesian generalization of standard Kalman filtering. To implement the theory we derive a parallel network which 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.