logo资料库

TinyML Machine Learning with TensorFlow Lite on Arduino and Micr....pdf

发布时间：2022-05-31 发布人：admin 分类：说明书 资料大小：23.43M 资料格式：pdf 举报 版权申诉
第1页 / 共504页
第2页 / 共504页
第3页 / 共504页
第4页 / 共504页
第5页 / 共504页
第6页 / 共504页
第7页 / 共504页
第8页 / 共504页
资料共504页，剩余部分请下载后查看
    Cover
    Copyright
    Table of Contents
    Preface
    Conventions Used in This Book
    Using Code Examples
    O’Reilly Online Learning
    How to Contact Us
    Acknowledgments
    Chapter 1. Introduction
    Embedded Devices
    Changing Landscape
    Chapter 2. Getting Started
    Who Is This Book Aimed At?
    What Hardware Do You Need?
    What Software Do You Need?
    What Do We Hope You’ll Learn?
    Chapter 3. Getting Up to Speed on Machine Learning
    What Machine Learning Actually Is
    The Deep Learning Workflow
    Decide on a Goal
    Collect a Dataset
    Design a Model Architecture
    Train the Model
    Convert the Model
    Run Inference
    Evaluate and Troubleshoot
    Wrapping Up
    Chapter 4. The “Hello World” of TinyML: Building and Training a Model
    What We’re Building
    Our Machine Learning Toolchain
    Python and Jupyter Notebooks
    Google Colaboratory
    TensorFlow and Keras
    Building Our Model
    Importing Dependencies
    Generating Data
    Splitting the Data
    Defining a Basic Model
    Training Our Model
    Training Metrics
    Graphing the History
    Improving Our Model
    Testing
    Converting the Model for TensorFlow Lite
    Converting to a C File
    Wrapping Up
    Chapter 5. The “Hello World” of TinyML: Building an Application
    Walking Through the Tests
    Including Dependencies
    Setting Up the Test
    Getting Ready to Log Data
    Mapping Our Model
    Creating an AllOpsResolver
    Defining a Tensor Arena
    Creating an Interpreter
    Inspecting the Input Tensor
    Running Inference on an Input
    Reading the Output
    Running the Tests
    Project File Structure
    Walking Through the Source
    Starting with main_functions.cc
    Handling Output with output_handler.cc
    Wrapping Up main_functions.cc
    Understanding main.cc
    Running Our Application
    Wrapping Up
    Chapter 6. The “Hello World” of TinyML: Deploying to Microcontrollers
    What Exactly Is a Microcontroller?
    Arduino
    Handling Output on Arduino
    Running the Example
    Making Your Own Changes
    SparkFun Edge
    Handling Output on SparkFun Edge
    Running the Example
    Testing the Program
    Viewing Debug Data
    Making Your Own Changes
    ST Microelectronics STM32F746G Discovery Kit
    Handling Output on STM32F746G
    Running the Example
    Making Your Own Changes
    Wrapping Up
    Chapter 7. Wake-Word Detection: Building an Application
    What We’re Building
    Application Architecture
    Introducing Our Model
    All the Moving Parts
    Walking Through the Tests
    The Basic Flow
    The Audio Provider
    The Feature Provider
    The Command Recognizer
    The Command Responder
    Listening for Wake Words
    Running Our Application
    Deploying to Microcontrollers
    Arduino
    SparkFun Edge
    ST Microelectronics STM32F746G Discovery Kit
    Wrapping Up
    Chapter 8. Wake-Word Detection: Training a Model
    Training Our New Model
    Training in Colab
    Using the Model in Our Project
    Replacing the Model
    Updating the Labels
    Updating command_responder.cc
    Other Ways to Run the Scripts
    How the Model Works
    Visualizing the Inputs
    How Does Feature Generation Work?
    Understanding the Model Architecture
    Understanding the Model Output
    Training with Your Own Data
    The Speech Commands Dataset
    Training on Your Own Dataset
    How to Record Your Own Audio
    Data Augmentation
    Model Architectures
    Wrapping Up
    Chapter 9. Person Detection: Building an Application
    What We’re Building
    Application Architecture
    Introducing Our Model
    All the Moving Parts
    Walking Through the Tests
    The Basic Flow
    The Image Provider
    The Detection Responder
    Detecting People
    Deploying to Microcontrollers
    Arduino
    SparkFun Edge
    Wrapping Up
    Chapter 10. Person Detection: Training a Model
    Picking a Machine
    Setting Up a Google Cloud Platform Instance
    Training Framework Choice
    Building the Dataset
    Training the Model
    TensorBoard
    Evaluating the Model
    Exporting the Model to TensorFlow Lite
    Exporting to a GraphDef Protobuf File
    Freezing the Weights
    Quantizing and Converting to TensorFlow Lite
    Converting to a C Source File
    Training for Other Categories
    Understanding the Architecture
    Wrapping Up
    Chapter 11. Magic Wand: Building an Application
    What We’re Building
    Application Architecture
    Introducing Our Model
    All the Moving Parts
    Walking Through the Tests
    The Basic Flow
    The Accelerometer Handler
    The Gesture Predictor
    The Output Handler
    Detecting Gestures
    Deploying to Microcontrollers
    Arduino
    SparkFun Edge
    Wrapping Up
    Chapter 12. Magic Wand: Training a Model
    Training a Model
    Training in Colab
    Other Ways to Run the Scripts
    How the Model Works
    Visualizing the Input
    Understanding the Model Architecture
    Training with Your Own Data
    Capturing Data
    Modifying the Training Scripts
    Training
    Using the New Model
    Wrapping Up
    Learning Machine Learning
    What’s Next
    Chapter 13. TensorFlow Lite for Microcontrollers
    What Is TensorFlow Lite for Microcontrollers?
    TensorFlow
    TensorFlow Lite
    TensorFlow Lite for Microcontrollers
    Requirements
    Why Is the Model Interpreted?
    Project Generation
    Build Systems
    Specializing Code
    Makefiles
    Writing Tests
    Supporting a New Hardware Platform
    Printing to a Log
    Implementing DebugLog()
    Running All the Targets
    Integrating with the Makefile Build
    Supporting a New IDE or Build System
    Integrating Code Changes Between Projects and Repositories
    Contributing Back to Open Source
    Supporting New Hardware Accelerators
    Understanding the File Format
    FlatBuffers
    Porting TensorFlow Lite Mobile Ops to Micro
    Separate the Reference Code
    Create a Micro Copy of the Operator
    Port the Test to the Micro Framework
    Build a Bazel Test
    Add Your Op to AllOpsResolver
    Build a Makefile Test
    Wrapping Up
    Chapter 14. Designing Your Own TinyML Applications
    The Design Process
    Do You Need a Microcontroller, or Would a Larger Device Work?
    Understanding What’s Possible
    Follow in Someone Else’s Footsteps
    Find Some Similar Models to Train
    Look at the Data
    Wizard of Oz-ing
    Get It Working on the Desktop First
    Chapter 15. Optimizing Latency
    First Make Sure It Matters
    Hardware Changes
    Model Improvements
    Estimating Model Latency
    How to Speed Up Your Model
    Quantization
    Product Design
    Code Optimizations
    Performance Profiling
    Optimizing Operations
    Look for Implementations That Are Already Optimized
    Write Your Own Optimized Implementation
    Taking Advantage of Hardware Features
    Accelerators and Coprocessors
    Contributing Back to Open Source
    Wrapping Up
    Chapter 16. Optimizing Energy Usage
    Developing Intuition
    Typical Component Power Usage
    Hardware Choice
    Measuring Real Power Usage
    Estimating Power Usage for a Model
    Improving Power Usage
    Duty Cycling
    Cascading Design
    Wrapping Up
    Chapter 17. Optimizing Model and Binary Size
    Understanding Your System’s Limits
    Estimating Memory Usage
    Flash Usage
    RAM Usage
    Ballpark Figures for Model Accuracy and Size on Different Problems
    Speech Wake-Word Model
    Accelerometer Predictive Maintenance Model
    Person Presence Detection
    Model Choice
    Reducing the Size of Your Executable
    Measuring Code Size
    How Much Space Is Tensorflow Lite for Microcontrollers Taking?
    OpResolver
    Understanding the Size of Individual Functions
    Framework Constants
    Truly Tiny Models
    Wrapping Up
    Chapter 18. Debugging
    Accuracy Loss Between Training and Deployment
    Preprocessing Differences
    Debugging Preprocessing
    On-Device Evaluation
    Numerical Differences
    Are the Differences a Problem?
    Establish a Metric
    Compare Against a Baseline
    Swap Out Implementations
    Mysterious Crashes and Hangs
    Desktop Debugging
    Log Tracing
    Shotgun Debugging
    Memory Corruption
    Wrapping Up
    Chapter 19. Porting Models from TensorFlow to TensorFlow Lite
    Understand What Ops Are Needed
    Look at Existing Op Coverage in Tensorflow Lite
    Move Preprocessing and Postprocessing into Application Code
    Implement Required Ops if Necessary
    Optimize Ops
    Wrapping Up
    Chapter 20. Privacy, Security, and Deployment
    Privacy
    The Privacy Design Document
    Using a PDD
    Security
    Protecting Models
    Deployment
    Moving from a Development Board to a Product
    Wrapping Up
    Chapter 21. Learning More
    The TinyML Foundation
    SIG Micro
    The TensorFlow Website
    Other Frameworks
    Twitter
    Friends of TinyML
    Wrapping Up
    Appendix A. Using and Generating an Arduino Library Zip
    Appendix B. Capturing Audio on Arduino
    Index
    About the Authors
    Colophon
    TinyML Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers Pete Warden & Daniel Situnayake
    TinyML Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers Pete Warden and Daniel Situnayake Beijing Beijing Boston Boston Farnham Sebastopol Farnham Sebastopol Tokyo Tokyo
    TinyML by Pete Warden and Daniel Situnayake Copyright © 2020 Pete Warden and Daniel Situnayake. All rights reserved. Printed in the United States of America. Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA 95472. O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are also available for most titles (http://oreilly.com). For more information, contact our corporate/institutional sales department: 800-998-9938 or corporate@oreilly.com. Acquisitions Editor: Mike Loukides Development Editor: Nicole Taché Production Editor: Beth Kelly Copyeditor: Octal Publishing, Inc. December 2019: First Edition Revision History for the First Edition 2019-12-13: First Release Proofreader: Rachel Head Indexer: WordCo, Inc. Interior Designer: David Futato Illustrator: Rebecca Demarest See http://oreilly.com/catalog/errata.csp?isbn=9781492052043 for release details. The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. TinyML, the cover image, and related trade dress are trademarks of O’Reilly Media, Inc. TinyML is a trademark of the tinyML Foundation and is used with permission. While the publisher and the authors have used good faith efforts to ensure that the information and instructions contained in this work are accurate, the publisher and the authors disclaim all responsibility for errors or omissions, including without limitation responsibility for damages resulting from the use of or reliance on this work. Use of the information and instructions contained in this work is at your own risk. If any code samples or other technology this work contains or describes is subject to open source licenses or the intellectual property rights of others, it is your responsibility to ensure that your use thereof complies with such licenses and/or rights. 978-1-492-05204-3 [LSI]
    Table of Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Embedded Devices 3 Changing Landscape 4 2. Getting Started. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Who Is This Book Aimed At? 5 What Hardware Do You Need? 6 What Software Do You Need? 7 What Do We Hope You’ll Learn? 8 3. Getting Up to Speed on Machine Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 What Machine Learning Actually Is 12 The Deep Learning Workflow 13 Decide on a Goal 14 Collect a Dataset 14 Design a Model Architecture 16 Train the Model 21 Convert the Model 26 Run Inference 26 Evaluate and Troubleshoot 27 Wrapping Up 28 4. The “Hello World” of TinyML: Building and Training a Model. . . . . . . . . . . . . . . . . . . . . . 29 What We’re Building 30 Our Machine Learning Toolchain 32 Python and Jupyter Notebooks 32 iii
    Google Colaboratory 33 TensorFlow and Keras 33 Building Our Model 34 Importing Dependencies 35 Generating Data 38 Splitting the Data 41 Defining a Basic Model 42 Training Our Model 46 Training Metrics 48 Graphing the History 49 Improving Our Model 54 Testing 58 Converting the Model for TensorFlow Lite 60 Converting to a C File 64 Wrapping Up 65 5. The “Hello World” of TinyML: Building an Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Walking Through the Tests 68 Including Dependencies 69 Setting Up the Test 70 Getting Ready to Log Data 70 Mapping Our Model 72 Creating an AllOpsResolver 74 Defining a Tensor Arena 74 Creating an Interpreter 75 Inspecting the Input Tensor 75 Running Inference on an Input 78 Reading the Output 80 Running the Tests 82 Project File Structure 85 Walking Through the Source 86 Starting with main_functions.cc 87 Handling Output with output_handler.cc 90 Wrapping Up main_functions.cc 91 Understanding main.cc 91 Running Our Application 92 Wrapping Up 93 6. The “Hello World” of TinyML: Deploying to Microcontrollers. . . . . . . . . . . . . . . . . . . . . . 95 What Exactly Is a Microcontroller? 96 Arduino 97 Handling Output on Arduino 98 iv | Table of Contents
    Running the Example 101 Making Your Own Changes 106 SparkFun Edge 106 Handling Output on SparkFun Edge 107 Running the Example 110 Testing the Program 117 Viewing Debug Data 118 Making Your Own Changes 118 ST Microelectronics STM32F746G Discovery Kit 119 Handling Output on STM32F746G 119 Running the Example 124 Making Your Own Changes 126 Wrapping Up 126 7. Wake-Word Detection: Building an Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 What We’re Building 128 Application Architecture 129 Introducing Our Model 130 All the Moving Parts 132 Walking Through the Tests 133 The Basic Flow 134 The Audio Provider 138 The Feature Provider 139 The Command Recognizer 145 The Command Responder 151 Listening for Wake Words 152 Running Our Application 156 Deploying to Microcontrollers 156 Arduino 157 SparkFun Edge 165 ST Microelectronics STM32F746G Discovery Kit 175 Wrapping Up 180 8. Wake-Word Detection: Training a Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Training Our New Model 182 Training in Colab 182 Using the Model in Our Project 197 Replacing the Model 197 Updating the Labels 198 Updating command_responder.cc 198 Other Ways to Run the Scripts 201 How the Model Works 202 Table of Contents | v
    Visualizing the Inputs 202 How Does Feature Generation Work? 206 Understanding the Model Architecture 208 Understanding the Model Output 213 Training with Your Own Data 214 The Speech Commands Dataset 215 Training on Your Own Dataset 216 How to Record Your Own Audio 216 Data Augmentation 218 Model Architectures 218 Wrapping Up 219 9. Person Detection: Building an Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 What We’re Building 222 Application Architecture 224 Introducing Our Model 224 All the Moving Parts 225 Walking Through the Tests 227 The Basic Flow 227 The Image Provider 231 The Detection Responder 232 Detecting People 233 Deploying to Microcontrollers 235 Arduino 236 SparkFun Edge 246 Wrapping Up 257 10. Person Detection: Training a Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Picking a Machine 259 Setting Up a Google Cloud Platform Instance 260 Training Framework Choice 268 Building the Dataset 269 Training the Model 270 TensorBoard 272 Evaluating the Model 274 Exporting the Model to TensorFlow Lite 274 Exporting to a GraphDef Protobuf File 274 Freezing the Weights 275 Quantizing and Converting to TensorFlow Lite 275 Converting to a C Source File 276 Training for Other Categories 277 Understanding the Architecture 277 vi | Table of Contents
    分享到：
    赞 收藏

    相关推荐

    资料库

    人工智能

    共收录2047份资料，累计62个分类，关注成员有19位，主要包括：GNSS,Pico-SIM868-GSM,GPRS,NB-IoT,GPRSHAT,SIM7070GCat-M,GNSSHAT,GSM,NB-IoTHAT,SIM7080GCat-M,R800CGSM,SIM800CGSM,UPSPowerModule(B),UPSPowerModule,Pico-SIM7020C-NB-IoT,SIM7028NB-IoTHAT,SIM7020CNB-IoTHAT,SIM7000XNB-IoTHAT,A7670CCat-1HAT,A7600C1Cat-1,SIM7600X4G&LTECat-1HAT,SIM7600X4GDONGLE用户手册,RM520N-GL5GHAT,RM500Q-GL5GHAT,RM500U-CN5GHAT,SIM820X-M25GforJetsonNano,AudioCardforJetsonNano,SIM7600G-H4GforJetsonNano,OpenNCC-Lite-4G8M,IMX219-77Camera,13.3inche-PaperHAT,10.3inche-PaperHAT(D),10.3inche-PaperHAT,9.7inche-PaperHAT,7.8inche-PaperHAT,6inchHDTouche-PaperHAT,6inchHDe-PaperHAT,6inche-PaperHAT,5.83inche-PaperHAT(C),LCD-Clock-A,0.95inchRGBOLED(A),0.96inchRGBOLEDModule,0.96inchOLEDModule,0.96inchOLED(A),1.27inchRGBOLEDModule,1.3inchOLEDModule(C),1.3inchOLED(A),1.32inchOLEDModule,1.5inchRGBOLEDModule,1.5inchOLEDModule,JETSONORINNX,自然语言处理,搜索引擎,VR,计算广告,深度学习,机器学习

    热门标签

    GNSS Pico-SIM868-GSM GPRS NB-IoT GPRSHAT SIM7070GCat-M GNSSHAT GSM NB-IoTHAT SIM7080GCat-M R800CGSM SIM800CGSM UPSPowerModule(B) UPSPowerModule Pico-SIM7020C-NB-IoT SIM7028NB-IoTHAT SIM7020CNB-IoTHAT SIM7000XNB-IoTHAT A7670CCat-1HAT A7600C1Cat-1 SIM7600X4G&LTECat-1HAT SIM7600X4GDONGLE用户手册 RM520N-GL5GHAT RM500Q-GL5GHAT RM500U-CN5GHAT SIM820X-M25GforJetsonNano AudioCardforJetsonNano SIM7600G-H4GforJetsonNano OpenNCC-Lite-4G8M IMX219-77Camera 13.3inche-PaperHAT 10.3inche-PaperHAT(D) 10.3inche-PaperHAT 9.7inche-PaperHAT 7.8inche-PaperHAT 6inchHDTouche-PaperHAT 6inchHDe-PaperHAT 6inche-PaperHAT 5.83inche-PaperHAT(C) LCD-Clock-A 0.95inchRGBOLED(A) 0.96inchRGBOLEDModule 0.96inchOLEDModule 0.96inchOLED(A) 1.27inchRGBOLEDModule 1.3inchOLEDModule(C) 1.3inchOLED(A) 1.32inchOLEDModule 1.5inchRGBOLEDModule 1.5inchOLEDModule JETSONORINNX 自然语言处理 搜索引擎 VR 计算广告 深度学习 机器学习

    最新资料