Photo Content Classification Using Convolutional Neural Network

Abstract

This research is about photo content classification using the TensorFlow framework and the deep neural network algorithm in deep learning. Machine learning is a multidisciplinary major whose main research object is artificial intelligence, which enables machines to have the learning ability like human beings. As a branch of machine learning, deep learning analyses and interprets texts, images, and sounds by learning the rules and representations of sample data. A convolutional neural network is an approach of a deep neural network, which is mostly applied to analyse the photo content and has good performance in photo content classification. It is also the main point that the study focuses on. The Visual Geometry Group 16 (VGG16) model in the convolutional neural network was used to achieve the research objectives and has excellent performance. The research used data from CelebFaces and Kaggle, including photos about people, food, animals, and architecture. The photo content classification model in this study can classify these four types of images. Moreover, the analysis of the work, summary, as well as the future work were provided at the end.