Optimization of AI Models for Embedded Systems: A Literature Review and Comparative Analysis
Assia Belatik, Hasnae El Khoukhi, My Abdelouahed Sabri, Abdellah Aarab · 2025
In order to deploy complex machine learning models on resource-constrained devices such as microcontrollers, FPGAs, and edge devices, the optimization of artificial intelligence (AI) models for embedded systems is a crucial research area. Computational power, memory, and energy are the limitations these devices face that make it difficult to run AI models efficiently. This literature review offers an overview of various strategies and techniques used to optimize AI models for embedded systems, emphasizing on area like model compression, hardware acceleration, resource management, and co-design approaches. The review focuses on strategies and tools used to enhance the performance and efficiency of AI models in these environments.