Regresar

Machine learning for computer vision: A review of theory and algorithms

Abstract:

The Computer vision is a problem that belongs to the field of artificial intelligence. The problem of computer vision consists of designing algorithms that allow computers to simulate the ability of humans to see the world. The three main problems that computer vision deals with include image segmentation, image classification and object detection. In this paper, we present a brief and technical review of the theory and algorithms of machine learning that can be applied to these three problems of computer vision. Specifically, we present a formal definition of each of these three problems, review the theory of machine learning for designing classifiers under the context of random processes, and then describe the most popular algorithms of machine learning applied to computer vision including feed-forward artificial neural networks, convolutional neural networks and deep learning. Finally, we hope that the self-contained material of this paper provides to the readers with enough information to start using machine learning for solving certain problems of computer vision.