79,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
Comprehensive, accessible introduction to deep learning for engineering tasks through Python programming, low-cost hardware, and freely available software
Deep Learning on Embedded Systems is a comprehensive guide to the practical implementation of deep learning for engineering tasks through computers and embedded hardware such as Raspberry Pi and Nvidia Jetson Nano. After an introduction to the field, the book provides fundamental knowledge on deep learning, convolutional and recurrent neural networks, computer vision, and basics of Linux terminal and docker engines. This book shows detailed setup steps of Jetson Nano and Raspberry Pi for utilizing essential frameworks such as PyTorch and OpenCV. GPU configuration and dependency installation procedure for using PyTorch is also discussed allowing newcomers to seamlessly navigate the learning curve.
A key challenge of utilizing deep learning on embedded systems is managing limited GPU and memory resources. This book outlines a strategy of training complex models on a desktop computer and transferring them to embedded systems for inference. Also, students and researchers often face difficulties with the varying probabilistic theories and notations found in data science literature. To simplify this, the book mainly focuses on the practical implementation part of deep learning using Python programming, low-cost hardware, and freely available software such as Anaconda and Visual Studio Code. To aid in reader learning, questions and answers are included at the end of most chapters.
Written by a highly qualified author, Deep Learning on Embedded Systems includes discussion on:
- Fundamentals of deep learning, including neurons and layers, activation functions, network architectures, hyperparameter tuning, and convolutional and recurrent neural networks (CNNs & RNNs)
- PyTorch, OpenCV, and other essential framework setups for deep transfer learning, along with Linux terminal operations, docker engine, docker images, and virtual environments in embedded devices
- Training models for image classification and object detection with classification, then converting trained PyTorch models to ONNX format for efficient deployment on Jetson Nano and Raspberry Pi
Deep Learning on Embedded Systems serves as an excellent introduction to the field for undergraduate engineering students seeking to learn deep learning implementations for their senior capstone or class projects and graduate researchers and educators who wish to implement deep learning in their research.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 400
Veröffentlichungsjahr: 2025
Ähnliche
Table of Contents
Cover
Table of Contents
Title Page
Copyright
Preface
Acknowledgment
Biography
About the Companion Website
Chapter 1: Introduction
1.1 Machine Learning to Deep Learning
1.2 Modern Embedded Systems
1.3 Deep Transfer Learning in Embedded System
1.4 Deep Learning Frameworks: An Overview
1.5 Deep Learning and AI: Big Data and the Road Ahead
References
Chapter 2: Fundamentals of Deep Learning
2.1 Neural Networks Overview
2.2 Basic Concepts and Terminologies
2.3 Training on a Network
2.4 Gradient Descent Algorithm
2.5 Weight Initialization and Regularization
2.6 Hyperparameter Tuning
2.7 Overview of Common Hyperparameters
2.8 Challenges and Best Practices for Efficient Tuning
References
Chapter 3: Convolutional and Recurrent Neural Network
3.1 Introduction
3.2 Historical Background
3.3 Convolutional Neural Networks
3.4 Recurrent Neural Networks
3.5 Applications
3.6 Conclusion
References
Chapter 4: Deep Learning Using PyTorch
4.1 Introduction to PyTorch
4.2 Anaconda and PyTorch for Windows System
4.3 Other Essential Packages for Deep Learning
4.4 Introduction to Tensor
4.5 Basic Torch Operations in PyTorch
4.6 Gradient Calculation in PyTorch
4.7 Exercise Problems
References
Chapter 5: Introduction to Jetson Nano and Setup
5.1 Introduction to Jetson Embedded Devices
5.2 Jetpack Installation
5.3 Direct Setup
5.4 Configure Visual Studio Code on Jetson
5.5 OpenCV and PyTorch in Jetson
5.6 Setting up Jetson Inference
5.7 OpenCV Library and Test Video Capture Functionality
5.8 Conclusion
5.9 Exercise Problems
References
Chapter 6: Linux Terminal Overview
6.1 Introduction
6.2 Basic Terminal Commands and Syntax
6.3 Overview of File System
6.4 Navigating Files and Directories
6.5 Create, Edit, and Delete
6.6 Create and Execute Python Code from Terminal
6.7 Common Wildcard Characters
6.8 Find, View, and Get Information
6.9 Permission and Ownership
6.10 Install and Uninstall Packages Using “sudo”
6.11 Conclusion
References
Chapter 7: Docker Engine Setup
7.1 Introduction to Docker Engine
7.2 Docker in Embedded Devices
7.3 Jetson Inference Docker
7.4 Using Host Files in Docker Environment
7.5 Building a Docker Image
7.6 Run Python Through Docker Container
7.7 Exercise Problems
References
Chapter 8: Dataset Development
8.1 Introduction and Requirements
8.2 Types of Datasets
8.3 Manual Dataset Creation
8.4 Automatic Image Collection Using Embedded Device
8.5 Automatic Data Labeling
8.6 Data Preprocessing and Cleaning
8.7 Exercise Problem
References
Chapter 9: Training Model for Image Classification
9.1 Problem Statement
9.2 Default Configurations and Libraries
9.3 Setup Data Frame Using Annotations
9.4 Dataset Class and Methods
9.5 Data Loader and Model Configuration
9.6 Model Training
9.7 Testing and Inference
9.8 Fine-tuning
9.9 Application in Embedded System
9.10 Exercise Problems
References
Chapter 10: Object Detection with Classification
10.1 Introduction
10.2 Import Modules and Libraries
10.3 Default Configurations and Random Seeds
10.4 Create Data Frame and Process Labels
10.5 Training and Validation of Transformers
10.6 Dataset Class and Methods
10.7 Data Loader and Classification Backbone
10.8 Training and Validation Approach
10.9 Run Multiple Epochs and Save the Best
10.10 Model Inference
10.11 Multiple Object Detection with Classification
10.12 Model Inference for Multiple Object Detection
10.13 Conclusion
10.14 Exercise Problems
References
Chapter 11: Deploy Deep Learning Models on Jetson Nano
11.1 Introduction
11.2 Pre-trained Models
11.3 Inference on an Image File
11.4 ONNX Model
11.5 Inference on Live Video Stream
11.6 Conclusion
11.7 Exercise Problem
References
Chapter 12: Trained PyTorch Model: From Desktop PC to Jetson Nano
12.1 Introduction
12.2 Model Training on a PC
12.3 ONNX Model Inference
12.4 Conclusion
12.5 Exercise Problems
Chapter 13: Setting up Raspberry Pi
13.1 Introduction to Raspberry Pi
13.2 Hardware and Power Requirements
13.3 Operating System Setup
13.4 Create Virtual Environment
13.5 PyTorch and OpenCV Installation
13.6 Other Essential Packages
13.7 Conclusion
13.8 Exercise Problem
References
Chapter 14: Deploy Deep Learning Models on Raspberry Pi
14.1 Introduction
14.2 Face Detection and Recognition in Video Feeds
14.3 Real-time Object Detection
14.4 Real-time Classification
14.5 Real-time Segmentation
14.6 Exercise Problems
References
Chapter 15: Trained PyTorch Model: From Desktop PC to Raspberry Pi
15.1 Introduction
15.2 Model Training on a Desktop PC
15.3 PyTorch’s.pth to ONNX
15.4 ONNX Model Inference on Raspberry Pi 5
15.5 Conclusion
15.6 Exercise Problems
Reference
Index
End User License Agreement
List of Tables
Chapter 1
Table 1.1 Processing power and memory of common embedded devices.
Table 1.2 Deep learning framework, developer, and key features.
Chapter 5
Table 5.1 Descriptions of the hardware components used and alternatives.
Chapter 13
Table 13.1 Hardware components used for Raspberry Pi 5 setup.
Chapter 14
Table 14.1 Download the YOLOv3 model’s configuration, weights, and COCO label n...
Guide
Cover
Table of Contents
Title Page
Copyright
Preface
Acknowledgment
Biography
About the Companion Website
Begin Reading
Index
End User License Agreement
Pages
iii
iv
x
xi
xii
xiii
xiv
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
Deep Learning on Embedded Systems
A Hands-On Approach Using Jetson Nano and Raspberry Pi
By Tariq M. Arif
Copyright © 2025 by John Wiley & Sons Inc. All rights reserved, including rights for text and data mining and training of artificial intelligence technologies or similar technologies.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.
The manufacturer’s authorized representative according to the EU General Product Safety Regulation is Wiley-VCH GmbH, Boschstr. 12, 69469 Weinheim, Germany, e-mail: [email protected].
Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.
Library of Congress Cataloging-in-Publication Data
Names: Arif, Tariq M., author. | John Wiley & Sons, publisher.
Title: Deep learning on embedded systems : a hands-on approach using Jetson Nano and Raspberry Pi / by Tariq M. Arif.
Description: Hoboken, New Jersey : Wiley, [2025] | Includes bibliographical references and index. | Summary: “Preface The field of Artificial intelligence has undergone a significant transformation in the last decade, moving from traditional machine learning approaches to more sophisticated deep learning techniques. This evolution has brought extraordinary advancements across various industries, including healthcare, finance, transportation, manufacturing, robotics, and consumer technology. For this reason, there is a growing need to incorporate deep learning technology in various research projects and academic curricula. As customizable embedded devices become more affordable and portable for deploying AI models, the growing demand for exploring this technology is also spreading across all age groups, from children to the elderly. This book aims to address this demand and serves as a comprehensive hands-on guide to understanding the integration of deep learning with modern embedded systems, such as Jetson Nano and Raspberry Pi. It also focuses on the key components of deep learning models in simple terms without diving deeply into the statistical or mathematical theories behind them. A basic understanding of Python programming is necessary to follow the examples, as all the programs in this book are written in Python. The book introduces key concepts of deep learning and its architectures in chapters 2 and 3. Chapter 4 includes the configuration of the Windows PC used for setting up PyTorch and its related packages. This chapter also explains basic tensor operations using PyTorch. Chapter 5 and Chapter 13 include Jetson Nano and Raspberry Pi 5 configurations, respectively, along with the list of peripherals used for deploying deep learning models. As the operation of Jetson Nano and Raspberry Pi 5 involves using Linux terminals, Chapter 6 covers basic Linux terminal commands, focusing on file management and permissions. This chapter will be beneficial for readers who are unfamiliar with the Linux operating system. Chapter 7 presents the fundamentals of setting up the Docker engines and building Docker images, and demonstrates how to perform model inference within Jetson’s Docker container. Chapter 11 explains how to create a deep-learning dataset for image classification and object detection using bounding boxes. The dataset developed in this chapter is utilized for model training in Chapters 9 and 10. Chapter 9 outlines the process for training a classification model, while Chapter 10 demonstrates the approach for training an object detection model with image classification”– Provided by publisher.
Identifiers: LCCN 2025007090 | ISBN 9781394269266 (hardback) | ISBN 9781394269280 (adobe pdf) | ISBN 9781394269273 (epub) | ISBN 9781394269297 (ebook other)
Subjects: LCSH: Deep learning (Machine learning). | Microcomputers–Programming. | Embedded computer systems–Programming. | Image processing–Digital techniques.
Classification: LCC Q325.73 .A75 2025 | DDC 006.3/12–dc23/eng/20250319
LC record available at https://lccn.loc.gov/2025007090
Cover Design: Wiley
Cover Image: © Motion Loop/Shutterstock
Preface
The field of artificial intelligence has undergone a significant transformation in the last decade, moving from traditional machine learning approaches to more sophisticated deep learning techniques. This evolution has brought extraordinary advancements across various industries, including healthcare, finance, transportation, manufacturing, robotics, and consumer technology. For this reason, there is a growing need to incorporate deep learning technology in various research projects and academic curricula. As customizable embedded devices become more affordable and portable for deploying AI models, the growing demand for exploring this technology is also spreading across all age groups, from children to the elderly. This book aims to address this demand and serves as a comprehensive hands-on guide to understanding the integration of deep learning with modern embedded systems, such as Jetson Nano and Raspberry Pi. It also focuses on the key components of deep learning models in simple terms without diving deeply into the statistical or mathematical theories behind them.
A basic understanding of Python programming is necessary to follow the examples, as all the programs in this book are written in Python. The book introduces key concepts of deep learning and its architectures in Chapters 2 and 3. Chapter 4 includes the configuration of the Windows PC used for setting up PyTorch and its related packages. This chapter also explains basic tensor operations using PyTorch. Chapters 5 and 13 include Jetson Nano and Raspberry Pi 5 configurations, respectively, along with the list of peripherals used for deploying deep learning models. As the operation of the Jetson Nano and Raspberry Pi 5 involves using Linux terminals, Chapter 6 covers basic Linux terminal commands, focusing on file management and permissions. This chapter will be beneficial for readers who are unfamiliar with the Linux operating system. Chapter 7 presents the fundamentals of setting up the docker engines and building docker images and demonstrates how to perform model inference within Jetson’s docker container. Chapter 8 explains how to create a deep learning dataset for image classification and object detection using bounding boxes. The dataset developed in this chapter is utilized for model training in Chapters 9 and 10. Chapter 9 outlines the process for training a classification model, while Chapter 10 demonstrates the approach for training an object detection model with image classification.
A common challenge in deploying deep learning models on embedded systems is the setup process, especially when handling memory-intensive models. These devices often lack the memory capacity needed to train large deep learning models. To address this, in Chapters 12 and 15, the book introduces a method where models trained on a desktop computer can be transferred to Jetson Nano and Raspberry Pi for inference. This technique can be very useful for students working on science and engineering projects that involve deploying AI models on embedded devices.
Although there is abundant information available online, integrating deep learning models into design or research projects remains a challenging task. This book aims to be a practical guide for learning applied deep learning on embedded systems using Python. We encourage readers to follow the programming steps to reproduce the results demonstrated in the book before trying out the exercises. The hands-on approach presented here should provide a comprehensive understanding of the workflow, along with the related concepts and techniques.
Acknowledgment
I would like to express sincere thanks to Dr. Md Adilur Rahim from Louisiana State University for his contributions to earlier books, which laid the foundation for writing this one. I would also like to thank Devin Bigelow and Dalton Newbrough for their help in creating the datasets.
I need to appreciate my daughters, Nuha and Nesa, for their indirect contributions that made my work easy, especially when they thoughtfully chose not to interfere with my computer and toy-looking devices. Lastly, I owe special thanks to my wife, Mahbuba Sultana, whose patience and unwavering support made this work possible.
Biography
Tariq M. Arif is an associate professor in the Department of Mechanical Engineering at Weber State University, Utah. Before this role, he served as a lecturer faculty at the University of Wisconsin, Platteville. He earned his PhD in Mechanical Engineering from the New Jersey Institute of Technology (NJIT) in 2017. His primary research interests include artificial intelligence and genetic algorithms for robotics control, computer vision, and biomedical simulations involving machine learning algorithms. He obtained his master’s degree in 2011 from the University of Tokushima, Japan, and his BSc in 2005 from the Bangladesh University of Engineering and Technology (BUET). Tariq also worked in the Japanese automobile industry as a CAD/CAE engineer after completing his BSc degree. Throughout his industrial and academic career, he has participated in numerous research projects and authored several books on deep learning and its engineering applications. Currently, Tariq is working on the implementation of deep learning models for various computer vision-based controls and robotics applications.
About the Companion Website
This book is accompanied by a companion website:
www.wiley.com/go/Arif/Deep
This website includes:
Instructor
Student
Chapter 1Introduction
1.1 Machine Learning to Deep Learning
Machine learning is one of the artificial intelligence (AI) tools that emphasizes developing computer algorithms and statistical models to learn from data. It originated from computational learning theory and pattern recognition within AI. Machine learning algorithms create mathematical models using sample data (training data) to perform tasks without being explicitly programmed. To model complex patterns and relationships from training data, one can use neural networks, which consist of interconnected layers of nodes or neurons. This process of learning using neural networks emulates the information processing methods of human brains, which is why it is also referred to as artificial intelligence or AI. Over the last two decades, as the training data volume and complexity increased, traditional machine learning methods such as decision trees and support vector machines struggled with large-scale unstructured data such as images and text. This limitation led to the emergence of deep learning, a more advanced form of machine learning that utilizes many hidden layers in a neural network (typically 5–100) to analyze vast amounts of data and extract features. This sophisticated deep neural network approach, inspired by the human brain, has recently seen tremendous progress and accomplished significant milestones that would have been deemed impossible just a few decades ago.
The growth in deep learning technology can be attributed to several key factors: the exponential increase in computational power through the use of graphical processing units (GPUs) and tensor processing units (TPUs), advancements in neural network architectures, and the development of user-friendly, open-source deep learning frameworks such as TensorFlow and PyTorch. This field has become more accessible to researchers and practitioners, and substantial investments in this field from both academia and industry are helping continuous innovations in similar fields such as computer vision, natural language processing, and autonomous systems. As deep learning is reaching a more advanced stage of development, various scientific and engineering disciplines are recognizing its practical use cases. The emerging fields of robotics, AI, finance, environmental science, bioinformatics, genetics, and healthcare are increasingly adopting deep learning models, acknowledging their potential to drive significant advancements in their respective domains. This growing integration across diverse areas of study highlights the versatility and far-reaching impact of deep learning technology.
1.2 Modern Embedded Systems
In the last few years, the accessibility of powerful electronic processors has led to a surge in the incorporation of deep learning models into smaller devices. Also, as the electronic devices required for edge computing are getting cheaper and more accessible, there is a growing interest in using AI applications in embedded systems such as NVIDIA Jetson devices, Google Coral, Luxonis OAK-D cameras, Intel Movidius, etc. [1–4]. Deep learning enables these systems to perform complex tasks such as image and speech recognition, natural language processing, and anomaly detection, which were previously limited to larger, more powerful systems. Nowadays, many other embedded devices can process data locally and make real-time predictions and decisions, which is essential for autonomous driving, Internet of Things (IoT), robotics, and smart home devices.
Jetson’s embedded device is utilized for a wide range of applications in industry due to its powerful AI processing capabilities in a compact space. In manufacturing, it is commonly used in computer vision tasks, such as quality control and defect detection, which can enable real-time production line monitoring and analysis [5–7]. In robotics, it supports autonomous navigation and manipulation, which is crucial for improving the efficiency and accuracy of automated systems in warehouses and logistics [8, 9]. The Jetson devices play a wide variety of roles in IoT applications, including smart city infrastructure, where it helps manage traffic flow, monitor environmental conditions, and enhance public safety [10–13]. It is also employed in numerous other areas, such as retail for customer behavior analysis and inventory management and agriculture for crop monitoring and precision farming [7, 14–16]. In academia, embedded devices such as the Jetson Nano serve as a key tool in research and development, providing practical platforms for AI, machine learning, computer vision, and robotics experiments. Recently, these devices have been continuously integrated into AI courses, robotics and mechatronics programs, and IoT based lab applications, offering hands-on experience and enabling student capstone projects [17, 18]. They are also used extensively in hackathons and robotics competitions, fostering innovation and interdisciplinary collaborations. Overall, the combination of deep learning and embedded systems opens up numerous opportunities across various science and engineering domains by making advanced AI-driven functionalities widely accessible.
Currently, several embedded AI devices are utilized in industry and academia due to their powerful processing power, compact design, and real-time data analysis capabilities. These include the NVIDIA Jetson series, such as Jetson Nano, Jetson Xavier NX, and Jetson AGX Xavier, which are popular for computer vision, robotics, and IoT applications [19]. Google Coral devices, such as the Coral Dev Board and Coral USB Accelerator, are popular for prototyping and deploying machine learning models [20]. The Intel Neural Compute Stick 2 (discontinued after 2023) is known for AI capabilities in IoT devices, and the Raspberry Pi, often paired with AI accelerators such as the Coral USB Accelerator, is widely used in educational labs, projects, and small-scale AI experiments [21, 22]. A list of common embedded devices that can be utilized for small-scale AI applications and deep learning inference is summarized in Table 1.1.
Table 1.1 Processing power and memory of common embedded devices.
Embedded device
Configuration: AI performance, processing power, and memory
NVIDIA Jetson series
Jetson Nano
472 GFLOPs, Advanced RISC Machine (ARM) Cortex-A57 CPU, and a 128-core Maxwell GPU
2 GB or 4 GB LPDDR4 memory
(https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-nano/product-development/)
Jetson TX2
1.33 TFLOPs, Dual-core Denver 2 + Quad-core ARM Cortex-A57 CPU and 256-core Pascal GPU
32 GB eMMC 5.1 memory
(https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-tx2/)
Jetson Xavier NX
21 TOPs, 6-core NVIDIA Carmel ARM v8.2 64-bit CPU, and 384-core Volta with 48 Tensor Cores GPU
8 GB or 16 GB LPDDR4x memory
(https://www.arrow.com/en/products/900-83668-0000-000/nvidia)
Jetson AGX Xavier
32 TOPs, 8-core ARM v8.2 64-bit CPU, 8 MB L2 + 4 MB L3, and 512-core Volta with 64 Tensor Cores GPU
32 GB LPDDR4x memory and 32 GB eMMC 5.1 storage
(https://www.arrow.com/en/products/900-83668-0000-000/nvidia)
Other Jetson devices
Other variations of Jetson devices are available. A comprehensive list of specifications for these edge computing devices is available on their website.
(https://marketplace.nvidia.com/en-us/robotics-edge/)
Google coral devices
Coral Dev Board
4 TOPs, Integrated GC7000 Lite Graphics
NXP i.MX 8M system-on-chip (SoC) with a quad-core ARM Cortex-A53 CPU
1 or 4 GB LPDDR4 memory
(https://coral.ai/products/dev-board/#tech-specs)
Coral USB accelerator
Google Edge TPU coprocessor: 4 TOPS (int8), 2 TOPS per watt
It connects to an embedded system via USB 3.0 to provide AI acceleration.
(https://coral.ai/products/accelerator/)
Coral Dev Board Mini
4 TOPs, IMG PowerVR GE8300 (integrated in SoC)
MediaTek 8167s SoC (Quad-core Arm Cortex-A35) CPU
2 GB LPDDR3 memory and 8 GB eMMC flash memory.
(https://coral.ai/products/dev-board-mini/#tech-specs)
Coral system-on-module (SoM)
4 TOPs, Integrated GC7000 Lite Graphics
NXP i.MX 8M SoC (quad Cortex-A53, Cortex-M4F) CPU
1 GB LPDDR3 memory and 8 GB eMMC flash memory.
(https://coral.ai/products/som/#tech-specs)
Intel Compute Stick
Intel Neural Compute Stick 2
1 TOPs, Intel Movidius Myriad X VPU (Vision Processing Unit)
4 GB LPDDR4
Supports a wider range of AI frameworks through the OpenVINO toolkit. However, it is currently being phased out. (https://www.intel.com/content/www/us/en/products/sku/125743/intel-movidius-neural-compute-stick/specifications.html)
Raspberry Pi Boards
Raspberry Pi 5
VideoCore VII GPU, supporting OpenGL ES 3.1, Vulkan 1.2
Broadcom BCM2712 2.4 GHz quad-core 64-bit Arm Cortex-A76 CPU
LPDDR4X-4267 memory (4 GB or 8 GB)
Widely used in educational projects and small-scale AI experiments Can be paired with AI accelerators such as the Coral USB Accelerator.
(https://www.raspberrypi.com/products/raspberry-pi-5/)
Raspberry Pi 4 Model B
Broadcom BCM2711, quad-core Cortex-A72 (ARM v8) 64-bit CPU LPDDR4 memory (1 GB, 2 GB, 4 GB, or 8 GB)
Widely used in educational projects and small-scale AI experiments Can be paired with AI accelerators such as the Coral USB Accelerator.
(https://www.raspberrypi.com/products/raspberry-pi-4-model-b/)
In addition to these commonly used embedded devices, various other boards are capable of performing AI operations. For example, Texas Instrument’s Edge AI processor TI TDA4VM is utilized in automotive and industrial applications, and ARM Cortex-M Series processors, such as the Cortex-M7, are suitable for IoT sensors and low-power devices [23, 24]. Sony’s Spresense board can
