View dependence in scene recognition after active learning

Human spatial encoding of three-dimensional navigable space was studied, us ing a virtual environment simulation. This allowed subjects to become famil iar with a realistic scene by making simulated rotational and translational movements during training. Subsequent tests determined whether subjects co uld generalize their recognition ability by identifying novel-perspective v iews and topographic floor plans of the scene. Results from picture recogni tion tests showed that familiar direction views were most easily recognized , although significant generalization to novel views was observed. Topograp hic floor plans were also easily identified. In further experiments, novel- view performance diminished when active training was replaced by passive vi ewing of static images of the scene. However, the ability to make self-init iated movements, as opposed to watching dynamic movie sequences, had no eff ect on performance. These results suggest that representation of navigable space is view dependent and highlight the importance of spatial-temporal co ntinuity during learning.