Edge-based Cognitive System
Abstract
We investigate how to efficiently utilize the Edge computing resources to deploy the latency-sensitive applications in order to improve their Quality of Service (QoS) (e.g., latency). To this end, in the first part of our research, we want to develop edge-based applications, in particular, the cognitive assistance applications that in turn becoming the use case for our experiment and evaluation. The application we are building is assistance for assembling a toy car, in which one of the important tasks is the object detection module. We need to train the model for a highly accurate detection of different components of the car, which requires to have GPU resources.