Neural net simulations of human event parsing are described. A recurrent net was used to simulate data collected from human subjects watching short videotaped event sequences. In one simulation, the net was trained on one-half of a taped sequence with the other half of the sequence being used to test transfer performance. In another simulation, the net was trained on one complete event sequence and transfer to a different event sequence was tested. Neural net simulations provide a unique means of observing the interrelation of top-down and bottom-up processing in a basic cognitive task. Examination of computational patterns of the net and cluster analysis of the hidden units revealed two factors that may be central to event perception: (1) similarity between a current input and an activated schema and (2) expected duration of a given event. Although the importance of similarity between input and activated schemata during event perception has been acknowledged previously (e.g., Neisser, 1976; Schank, 1982), the present study provides specific instantiation of how similarity judgments can be made using both top-down and bottom-up processing. Moreover, unlike other work on event perception, this approach provides a potential mechanism for how schemata develop.

This content is only available as a PDF.