Natural Scene Statistics
Vision is the analysis of the visual patterns with a view to
understanding the objects and processes in the world that generate them.
There are a large number of visual areas in the cortex, devoted to
a variety of specific visual tasks, such as surface perception,
depth perception, motion analysis, object recognition.
Understanding how the brain accomplishes these tasks
requires an understanding of the statistical regularities of
visual patterns in the natural scenes and their relationship
to the desired results of the specific visual computations.
We have constructed a co-registered optical and range database
and are performing statistical analysis to gain a better
understanding of the neural representation and inference of 3D scenes.
We are also isolating statistical features in natural scenes
that are driving the neuronal adaptation mechanisms.
- Project 1: Computation and statistical analysis of 3D visual scenes
- Project Leader: Brian Potetz
- Project 2: Neurophysiology on the sensitivity of V1
neurons to statistical context
|
