Convolutional Neural Networks-An Extensive arena of Deep Learning. A Comprehensive Study: Fid-Bau Portal

Convolutional Neural Networks-An Extensive arena of Deep Learning. A Comprehensive Study

Singh, Navdeep / Sabrol, Hiteshwari

Abstract Deep learning is an evolving expanse of machine learning. Machine learning is observing its neoteric span as deep learning is steadily becoming the pioneer in this field. With the emergence of massive and wide-range technologies, recent advances in the area of deep learning have witnessed tremendous growth. Deep learning constitutes a subset of the algorithms of machine learning which aim to detect multiple level dispersion.In wide datasets the deep learning methodologies introduce non-linear transformations and high quality model abstractions. Convolutional neural networks (CNN), a benchmark in deep learning algorithms, have entirely transformed our understanding of the representation of information. CNN has the most reliable outcomes in the resolution of real world problems. In this article we offer a thorough description of CNN implementations in several application areas. On the basis of distinct research questions, we performed a systematic quantitative and performance analysis of the articles selected for this study. The majorly included segments for this study comprises of image classification, face recognition, human activity recognition, natural language processing and traffic management. Along with these, few other areas such as medical science, mechanics, social networking, computer networks and object detection are also addressed marginally. We also contrast the performance of CNN with diverse approaches and experience its effectiveness. This review article puts a limelight on CNN’s modern paradigm, its growing benefits, current implementations and success rate by illuminating on its various research developments.