Artificial Intelligence For Vision, An Overview
Abstract - We are in a modern period of digital time, where we have more digital devices than humans and its growing way too faster. At the same time, our demands are increasing exponentially. However, we are very much sure, No-one or individual organizations are not able to control this digital stream flow, and we will not be able to handle all of it by using our current capacity or resources as well. So we have (had) to updates generation after generation where digital technologies (AI) helped us to deal with these new revolutionary changes and still now, we are running for it. The point here, It’s not what the AI can do for us but the point here is what we can do with the AI. For the time being, still now it’s a mysterious box for us and we are still in a basic state with the AI. The technologies are tracing us and forcing us to resolve problems, again and again, to stay close to them but, we are not thinking onward or alternative manner, how to use them to make things in a different way before coming to an end of capacity or limitation rather than we have to think onward. In this paper, we will explore those development pathways that forward to best utilize computer vision and transfer learning to offer a whole bunch of ideas and inspirations to be captivated for thinking out of the box compare with future knowledge and avoid to involve technological deadlock.
Keywords - Artificial Intelligence, Technology Deadlock, Transfer Learning, Computer Vision, Federated Learning