A Survey Of Various Convolutional Neural Networks For Image ISSN: Recognition
AbstractDeep learning is a subset of Machine Learning which uses neural networks to mimic the workings of the human brain with the objective of analysing the data and identifying trends. Its usage in various fields like face recognition, speech to text conversion, has induced an exigency for constant updates and improvements to make it as reliable as a human being. This study presents a thorough examination of various Deep Learning approaches for Image Recognition by highlighting the challenges and contributions shown in neoteric research papers. Firstly, it gives a detailed overview of Neural Networks, their functioning, classifications and sub-classifications; and respective applications in diverse fields. It primarily analogizes one Deep Convolutional Neural Network architecture to another and provides a brief survey on the same. In closing, it offers a discourse on the challenges and future trends in training and designing deep neural networks.