Vision
- CABot1, 2 and 3 used a vision system based on a retina
of on-off and off-on center-surround receptors of varying sizes.
- They also had primary visual cortex which implemented, line, edge
and angle detectors.
- These are all reasonable models of biological vision.
- In CABot3, grating cells were implmented to enable texture
recognition. Again, a reasonable model of biological vision.
- Higher visual areas were not implmented, instead being replaced
by simple object recognizers that were not biologically
plausible.
- These recongized pyramids and stalactites in all three CABots.
- In CABot3, it also recognized corridors and door jams for
navigation.
