Evaluation Methods and Applications in Image Recognition based on Convolutional Neural Networks
Yiqi Christina Lin · Journal of Physics Conference Series · 2018
With the development of the technology in computer science and relative subjects, Machine Learning (ML) plays more and more important role in practical implementations. One of the most promising directions is the image recognition. This paper focuses on Convolutional Neural Networks (CNNs) which is the most common method in image recognition or pattern recognition issues. The basic understanding of CNNs' structure will be introduced, including the introduction of different layers. Also, two possible evaluation methods, based on NORB and MNIST respectively, will be displayed. The last part is some applications, followed by a summary.