SCENE DETECTION USING CONVOLUTION NEURAL NETWORK (CNN)

Abstract

Image classification is discipline of computer vision that deals with identifying and classifying objects from computer image based on certain properties. Image classification has a lot of applications such as robotics, smart traffic system, smart transportation and many more. The exponential growth of data in size and diversity cause major issue for image classification. Though, to detect and classify image from large dataset is require effective and efficient method of image classification. Therefore, machine learning algorithms are came into picture to achieve object detection. In this paper, scene detection is accomplished using machine learning algorithm named as convolution neural network (CNN). Scene detection is accomplished on Places Dataset. Places dataset is a repository of 10 million scene photographs, labeled with scene semantic categories, comprising a large and diverse list of the types of environments encountered in the world. The CNNs trained on Places shows that object detectors emerge as an intermediate representation of scene classification. With its high-coverage and high-diversity of examples, the Places Database along with the Places-CNNs offers a novel resource to guide future progress on scene recognition problems.