作者: Lucian-Ovidiu Fedorovici , Radu-Emil Precup , Florin Dragan , Radu-Codrut David , Constantin Purcaru
DOI: 10.1109/SACI.2012.6249989
关键词: Maxima and minima 、 Benchmark (computing) 、 Convolutional neural network 、 Process (computing) 、 Search algorithm 、 Computer science 、 Optical character recognition 、 Backpropagation 、 Embedding 、 Algorithm 、 Artificial intelligence
摘要: This paper presents aspects concerning embedding Gravitational Search Algorithms (GSAs) in Convolutional Neural Networks (CNNs) for Optical Character Recognition (OCR) systems. The GSAs are used combination with the Back Propagation (BP) algorithm as optimization algorithms training process of a specific CNN architecture OCR applications. new consists applying first GSA and next BP order to ensure performance improvements by avoiding algorithms' traps local minima. A analysis given benchmark application shows advantages our over classical six layer dedicated