Tiny Machine Learning (1 ed)

In the book Tiny Machine Learning: Design Principles and Applications, a distinguished team of researchers presents a comprehensive and structured exploration of the fundamental concepts, design principles, practical applications, and key challenges associated with Tiny Machine Learning (TinyML). The contributors introduce an innovative perspective on ultra-low-power computing resources, enabling a wide range of applications in Internet of Things (IoT) devices through system–algorithm co-design.

The book examines TinyML paradigms and enabling technologies, with a focus on applications such as anomaly detection and the broader learning landscape within resource-constrained environments. It provides clear explanations of TinyML devices and development tools, along with in-depth discussions on power consumption and memory constraints in IoT microcontrollers. In addition, it highlights lightweight frameworks tailored for efficient TinyML deployment and explores techniques suited for real-time and environmental applications.

Further topics covered in the book include:

• A comprehensive overview of security and privacy techniques for TinyML devices, including the implementation of advanced and novel protection schemes.
• Insightful analysis of power and memory optimization in IoT microcontrollers, emphasizing ultra-low-power smart devices with embedded TinyML capabilities.
• Practical perspectives on TinyML research focused on microcontroller-based data extraction and synthesis.

This book serves as a valuable resource for industry professionals, academic researchers, scientists, and engineers. It is also highly beneficial for lecturers and graduate students interested in advancing their knowledge of machine learning in resource-constrained environments.