Executive Summary : | With the advent of 5G technology, algorithms to process images as a service to the consumer need a thorough relook. Given dense connectivity of devices in 5G with sub-millisecond latency, the expectations from image processing algorithms and the nature of related services will change. For example, object recognition or understanding the content of an image or a video in a 5G infrastructure may depend on the fusion of multiple images or videos sourced from heterogeneous sensors having different resolutions. In this project, PIs propose to unravel such image or video processing models in a 5G infrastructure with a specific focus to investigate image and video understanding algorithms. The objective of the project is to study the deployment of image processing related services in a networked environment. The strategy for deployment of services will be guided by the principles of reinforcement learning. |