About the Book

Overview

Welcome to this collaborative project initiated by the CS249r Tiny Machine Learning class at Harvard University. Our goal is to make this book a community resource that assists educators and learners in understanding TinyML. The book will be regularly updated to reflect new insights into TinyML and effective teaching methods.

Topics Explored

This book offers a comprehensive look at various aspects of embedded machine learning. The topics we delve into include:

  • Introduction and Overview of Embedded Machine Learning
  • Data Engineering Techniques
  • Frameworks for Embedded Machine Learning
  • Efficient Representation and Compression of Models
  • Performance Metrics and Benchmarking for Machine Learning Systems
  • Edge Learning
  • Hardware Acceleration Options: GPUs, TPUs, and FPGAs
  • Operational Aspects of Embedded Machine Learning
  • Security and Privacy in On-Device Machine Learning
  • Ethical Considerations in AI
  • Sustainability Concerns in Edge Computing
  • Generative AI in Edge Computing

By the time you finish this book, you’ll have a foundational understanding of machine learning and the Internet of Things. You’ll also learn about real-world applications of embedded machine learning systems and gain practical experience through project-based assignments.

Who Should Read This

This book is tailored for those new to the exciting field of tiny machine learning (TinyML). It starts with the basics of machine learning and embedded systems and progresses to more advanced topics relevant to the TinyML community and broader research areas. The book is particularly beneficial for:

  • Embedded Systems Engineers: For engineers in the embedded systems domain, this book serves as an excellent guide to TinyML, helping them create intelligent applications on resource-limited platforms.

  • Students in Computer Science and Electrical Engineering: This book is a useful resource for students studying computer science and electrical engineering. It introduces them to the methods, algorithms, and techniques used in TinyML, preparing them for real-world challenges in embedded machine learning.

  • Researchers and Academics: Those involved in machine learning, computer vision, and signal processing research will find this book insightful. It sheds light on the unique challenges of running machine learning algorithms on low-power, low-memory devices.

  • Industry Professionals: If you’re working in areas like IoT, robotics, wearable tech, or smart devices, this book will equip you with the knowledge you need to add machine learning features to your products.

Key Learning Outcomes

Readers will acquire skills in training and deploying deep neural network models on resource-limited microcontrollers, along with understanding the broader challenges involved in their design, development, and deployment. Specifically, you’ll learn about:

  • Foundational Concepts in Machine Learning
  • Fundamentals of Embedded AI
  • Hardware Platforms Suitable for Embedded AI
  • Techniques for Training Models for Embedded Systems
  • Strategies for Model Optimization
  • Real-world Applications of Embedded AI
  • Current Challenges and Future Trends in Embedded AI

Our aim is to make this book a comprehensive resource for anyone interested in developing intelligent applications on embedded systems. Upon completing the book, you’ll be well-equipped to design and implement your own machine learning-enabled projects.

Prerequisites for Readers

  • Basic Programming Skills: We recommend that you have some prior programming experience, ideally in Python. A grasp of variables, data types, and control structures will make it easier to engage with the book.

  • Some Machine Learning Knowledge: While not mandatory, a basic understanding of machine learning concepts will help you absorb the material more readily. If you’re new to the field, the book provides enough background information to get you up to speed.

  • Python Programming (Optional): If you’re familiar with Python, you’ll find it easier to engage with the coding sections of the book. Knowing libraries like NumPy, scikit-learn, and TensorFlow will be particularly helpful.

  • Willingness to Learn: The book is designed to be accessible to a broad audience, with varying levels of technical expertise. A willingness to challenge yourself and engage in practical exercises will help you get the most out of it.

  • Resource Availability: For the hands-on aspects, you’ll need a computer with Python and the relevant libraries installed. Optional access to an embedded development board or microcontroller will also be beneficial for experimenting with machine learning model deployment.

By meeting these prerequisites, you’ll be well-positioned to deepen your understanding of TinyML, engage in coding exercises, and even implement practical applications on embedded devices.