Anfonwch hwn fel neges destun: Learning invariant object recognition in the visual system with continuous transformations.