Deep convolutional Neural networks(CNN) has recognized much advances in recent years. Many CNN models have been proposed in few years ago which focused by first on improving accuracy, next minimize number of parameters using squeeze architecture, then CNN model adapted for embedded and mobile systems. But face the huge applications of CNN in computer vision, few papers discuss what is the theory behind building CNN models combining some components. In this paper, we present a survey of recent advances in CNN architecture design taking into account the three periods listed above.
Paper
Full text
Trends in deep convolutional neural Networks architectures: a review
Semantic Scholar · Computer Science · 2019
Abstract
Deep convolutional Neural networks(CNN) has recognized much advances in recent years. Many CNN models have been proposed in few years ago which focused by first on improving accuracy, next minimize number of parameters using squeeze architecture, then CNN model adapted for embedded and mobile systems. But face the huge applications of CNN in computer vision, few papers discuss what is the theory behind building CNN models combining some components. In this paper, we present a survey of recent advances in CNN architecture design taking into account the three periods listed above.