Publication

Abstract : Due to the distributed representation of data through multiple levels of abstraction, one of the most well-liked branches of machine learning recently is “deep learning” (DL). An enormous amount of data is used to power deep learning systems. As huge amounts of data are produced, data analysis becomes increasingly difficult. Traditional Machine Language algorithms struggle to function well when there is a lot of data. Text, images, and other types of data can all be analyzed using deep learning, but deep learning resolves this problem. Deep learning can be used to analyze any kind of data, including text, images, and more. Deep learning, the next stage of machine learning in the last few decades, has revolutionized the way computer vision interprets human-centric content such as images, videos, sounds, and audio. This chapter aims to explain the fundamental deep learning algorithms that drive computer vision applications in the context of healthcare 4.0. The commonly used deep learning algorithms that are best suited for images are Stacked De-noising Auto-encoders, Convolutional Neural Networks (CNN), deep belief networks, and deep Boltzmann machines. Computer vision analytics with CNNs has many use cases, including recognition image based, detection. This helps in many areas to detect tumors in medical images and recognize their type, or helps robots identify and navigate obstacles. Despite considerable developments in computer vision using deep learning, until now there has been no software library covering these methods in a unified manner.