Edge AI: A Review of Machine Learning Models for Resource-Constrained Devices

Abstract

Edge Artificial Intelligence (Edge AI) is an emerging paradigm that integrates AI capabilities into edge devices, enabling real-time data processing and decision-making at the source of data generation. This article provides a comprehensive review of machine learning models tailored for resource-constrained devices, which are pivotal in the deployment of Edge AI. We explore the challenges and opportunities associated with implementing machine learning models on edge devices, including computational limitations, memory constraints, and energy efficiency. The review covers a range of machine learning techniques, from traditional models to advanced deep learning architectures, and discusses their adaptation for edge environments. Furthermore, we present three detailed tables summarizing the performance metrics, resource requirements, and application scenarios of various machine learning models in Edge AI. The article concludes with future research directions and potential advancements in the field.