作者: Roman M Borisyuk , Yakov B Kazanovich
DOI: 10.1016/J.NEUNET.2004.03.005
关键词: Novelty detection 、 Artificial neural network 、 Object (computer science) 、 Visual perception 、 Computer science 、 Representation (mathematics) 、 Artificial intelligence 、 Feature extraction 、 Object detection 、 Computer vision
摘要: We develop a new oscillatory model that combines consecutive selection of objects and discrimination between familiar objects. The works with visual information fulfils the following operations: (1) separation different according to their spatial connectivity; (2) located in field into attention focus; (3) extraction features; (4) representation working memory; (5) novelty detection functioning is based on two main principles: synchronization oscillators through phase-locking resonant increase amplitudes if they work in-phase other oscillators. results computer simulation are described for stimuli representing printed words.