About this Book

Several chapters in this book are also part of the open book Machine Learning Systems, which we invite you to read.

“Edge AI Engineering: Hands-on with the Raspberry Pi” is an accessible, practical guide designed to empower readers with the knowledge and skills needed to implement artificial intelligence (DL and GenAI) at the edge using the Raspberry Pi platform. This book is part of the open-source Machine Learning Systems initiative, which aims to democratize AI education and application.

Key Features:

1. Practical Approach: Each chapter is built around hands-on projects demonstrating real-world applications of edge ML on Raspberry Pi.

2. Progressive Learning: The book starts with fundamental concepts and progresses to more advanced topics, ensuring a smooth learning curve for readers of various skill levels.

3. Raspberry Pi Focus: All examples and projects are optimized for various Raspberry Pi models, including the Pi Zero 2W, Pi 4, and Pi 5, highlighting each model’s unique capabilities.

4. Comprehensive Coverage: From image processing and computer vision to natural language processing and sensor data analysis, this book covers various ML (and GenAI) applications relevant to edge computing.

5. Open-Source Tools: We emphasize using open-source models, frameworks, and libraries, such as Edge Impulse Studio, TensorFlow Lite, OpenCV, PyTorch, Transformers, and Ollama, ensuring accessibility and continuity in your learning journey.

6. Resource Optimization: Learn techniques to optimize ML models for the constrained resources of edge devices, balancing performance with efficiency.

7. Deployment Ready: Gain insights into best practices for deploying and maintaining ML models on Raspberry Pi in production environments.

Prerequisites:

While this book is designed to be accessible to a broad audience, readers will benefit from:

• Basic familiarity with Python programming
• Fundamental understanding of machine learning concepts
• Experience with Raspberry Pi or similar single-board computers (helpful but not required)

Structure of the Book:

The book is divided into chapters, each focusing on a specific aspect of edge ML on Raspberry Pi. Every chapter includes:

• Theoretical background to understand the concepts
• Step-by-step tutorials for implementing ML models
• Practical projects that apply the learned techniques
• Tips for troubleshooting and optimizing performance
• Suggestions for further exploration and experimentation

What’s Inside

• Introduction to EdgeAI and TinyML: A foundational look at embedded machine learning, including the differences between traditional AI, cloud AI, and AI at the edge.
• Getting Started with the Raspberry Pi: Learn to set up your Raspberry Pi for EdgeAI projects, including installation, system setup, and required libraries.
• Hands-On Projects: Step-by-step guides on implementing popular machine learning applications, such as image classification, object detection, anomaly detection, and more, directly on Raspberry Pi.
• Large Language Models at the Edge (SLMs): Explores running large language models (LLMs) on edge devices like the Raspberry Pi. It covers setting up Ollama and Python to leverage these models for tasks such as text generation, summarization, and conversational AI, making powerful language AI accessible at the edge.
• Vision-Language Models: Focusing on deploying Florence-2, Microsoft’s state-of-the-art Vision-Language Model, for various computer vision tasks such as captioning, object detection, segmentation, and visual grounding.
• Physical Computing and IoT Integration: Explores the implementation of Small Language Models (SLMs) in IoT control systems, demonstrating the possibility of creating a monitoring and control system using edge AI. These models will be integrated with physical sensors and actuators, creating an intelligent IoT system capable of natural language interaction.

By the end of this book, you’ll have a solid foundation in implementing various ML applications on Raspberry Pi and the confidence to tackle your edge AI projects.