A survey of Convolutional Neural Networks —From software to hardware and the applications in measurement

Abstract The convolutional neural network is a subfield of artificial neural networks and has made great achievements in various domains over the past decade. The technique has been widely applied including computer vision, natural language processing, image analysis, etc. The field of measurement also achieves exciting progress with convolutional neural networks. This paper gives a survey of convolutional neural networks from software algorithms to hardware architectures perspective and the applications in measurement. The survey hopes to give contributions to the cross-research of artificial intelligent and measurement.