Text this: Top-down resolution of visual ambiguity - knowledge from the future or footprints from the past?