Learning spatio-temporal relational structures
Walter F. Bischof, Terry Caelli · Applied Artificial Intelligence · 2001
We introduce a rule-based approach for learning and recognition of complex actions in terms of spatio-temporal attributes of primitive event sequences. During learning, spatio-temporal decision trees are generated which satisfy relational constraints of the training data. The resulting rules are used to classify new dynamic pattern fragments, and general heuristic rules are used to combine classification evidences of different pattern fragments.