AI Training System

Artificial Intelligence Training System: AI-22624

1. Features
1) Artificial Intelligence (AI) is a field that focuses on solving cognitive problems, primarily related to human intelligence, such as learning, problem-solving, and pattern recognition.
2) AI-22624 uses RaspberryPi and Python to learn the use of Google Assistant voice recognition and image processing (visual recognition) by using the camera.
3) With voice recognition, users can practice various voice recognition applications through the built-in sensors and actuators.
4) Through visual recognition, users learn object detection and recognition from various images and learn how to control the device.
5) Users also learn how to detect and recognize images using OpenCV, which is widely used for visual recognition.
6) In addition, the Jupiter Lab in Windows environment will be used to learn the basics of AI machine learning such as Linear Regression, Logistic Regression, Softmax classification, CNN... etc using TensorFlow and Python.
7) Users also learn how to improve the recognition rate by applying various algorithms of machine learning to MINIST number recognition.
8) From our products to our open-source platforms, we ensure that the benefits of AI are accessible to everyone.

2. Components
Hardware platform 1 set
Program CD 1 ea
USB Camera 1 ea
Speaker & USB Microphone 1 ea
Micro USB Cable 1 ea
Jumper cable 1 ea
DC 12V/3A Adaptor 1 ea
Micro SD Card & Reader 1 ea
HDMI Cable 1 ea
USB Keyboard & Mouse 1 ea
7 " LCD 1 ea
Textbook 2 books

3. Training Contents
< book-1, AI Voice recognition and Machine learning >
Chap 1, Overview of Artificial Intelligence (Theory)
1.1 Introduction to AI (Artificial Intelligence)
1.2 History and types of AI
Chap 2, Artificial Intelligence Classification / Product Configuration and Features (Theory)
2.1 Learning and reasoning
2.2 Understanding languages
2.3 Visual recognition
2.4 Situation awareness
2.5 AI Product Configuration and Features
Chap 3, RaspberryPi development environment (Practice)
3.1 About RaspberryPi
3.2 Hardware Assemble
3.3 Operating system installation
3.4 Network settings
3.5 SAMBA settings
3.6 GPIO introduction
3.7 GPIO library installation
3.8 System development method
Chap 4, Voice recognition and Google Assistant (Practice)
Introduction of Google Assistant
Build Google assistant development environment
Chap 5, Voice recognition device (Practice)
Device control method through speech recognition
LED control practice
Switch (KEYPAD) control practice
Chap 6, Voice recognition device (Practice)
FAN control practice
TEXTLCD control practice
Buzzer control practice
Chap 7, Interlock with voice recognition sensor -1 (Practice)
Interlock with luminance sensor
Interlock with GAS sensor
Chap 8, Interlock with voice recognition sensor -2 (Practices)
Interlock with Temperature and Humidity Sensor (DHT11)
Interlock with Motion sensor (PIR)
Chap 9, Design of voice recognition system (Theory)
Fire monitoring system
Cooling control system
Lighting control system
Intrusion detection system
Chap 10, Overview of Machine learning (Practice)
Introduction of Machine learning
Basics of Machine learning
Chap 11, Machine learning development environment (Theory & Practice)
Introduction of TensorFlow
Construction of TensorFlow development environment
TensorFlow API
Chap 12, Linear regression (Theory & Practice)
Theory of Linear regression
Minimize the losses
Multivariable Linear Regression
Chap 13, Machine Learning Practice (Theory & Practice)
Generalization, Over Integration, and Learning Rate Theory
Verification exercise
Chap 14, Machine learning and characteristics (Theory & Practice)
Character extraction and data refining theory
Cross-Cultural Practice
Chap 15, Logistic regression (Theory & Practice)
Theory of Logistic regression
Minimize the losses
Chap 16, Overview of Internet of Things (IoT)
Chap 17, Design of Internet of Things (IoT)

< book-2, AI Vision and Machine learning >
Chap 1, Overview of Artificial Intelligence (Theory)
1.1 Introduction of AI (Artificial Intelligence)
1.2 History and types of AI
Chap 2, Artificial Intelligence Classification / Product Configuration and Features (Theory)
2.1 Learning and reasoning
2.2 Understanding languages
2.3 Visual recognition
2.4 Situation awareness
2.5 AI Product Configuration and Features
Chap 3, RaspberryPi development environment (Practice)
3.1 About RaspberryPi
3.2 Accessories of RaspberryPi
3.3 Hardware Assemble
3.4 Operating system installation
3.5 Network settings
3.6 SAMBA setting
3.7 GPIO introduction
3.8 GPIO library installation
3.9 AI development environment setup
Chap 4, Visual recognition and Google Assistant (Practice)
Introduction of Google Assistant
Build Google assistant development environment
Visual recognition devices
Chap 5, Visual Recognition-1 using Google Vision API
Label recognition
Text recognition
Chap 6, Visual Recognition-2 using Google Vision API
Human face recognition
Facial expression recognition
Landmark recognition
Chap 7, Visual Recognition-3 using Google Vision API
Image properties
Safe search properties
Chap 8, Visual Recognition-4 using Google Vision API
Hints for cut
Web recognition
Chap 9, Visual Recognition-5 using Google Vision API
Document text recognition
Logo recognition
Chap 10, TensorFlow development environment
Introduction to TensorFlow
Development environment setting for TensorFlow
Use of TensorFlow library
Implement TensorFlow algorithm / programming
Chap 11, Image Recognition-1 using Machine Learning
Softmax classification
MNIST number recognition
Chap 12, Image Recognition-2 using Machine Learning
CNN (Convolutional Neural Network)
MNIST number recognition using CNN
Chap 13, Image Recognition-3 using Machine Learning
Introduction of objects recognition
Image recognition
Chap 14, OpenCV development environment
Introduction of OpenCV
OpenCV development environment setting
Chap 15, Visual recognition-1 using OpenCV
Use the camera
Human face recognition

4. Specs
1) Hardware specs
(1) RaspberryPi3 Board
Broadcom BCM2837B0, Cortex-A53 (ARMv8) 64-bit SoC @ 1.4GHz
2.4GHz and 5GHz IEEE 802.11.b/g/n/ac wireless LAN
Bluetooth 4.2, BLE
Gigabit Ethernet over USB 2.0 (maximum throughput 300 Mbps)
1GB LPDDR2 SDRAM
Extended 40-pin GPIO header
Full-size HDMI
4 USB 2.0 ports
CSI camera port for connecting a Raspberry Pi camera
DSI display port for connecting a Raspberry Pi touch screen display
4-pole stereo output and composite video port
Micro SD port for loading your operating system and storing data
5V/2.5A DC power input
Power-over-Ethernet (PoE) support (requires separate PoE HAT)
(2) Base Interface
Raspberry Pi 3 Adapter Board: LEVEL Buffer I/O Connector
Raspberry Pi 3 40-pin GPIO header
Raspberry Pi 3 8-pin Analog Input Port
Baseboard power supply switch
MUX Switch (SW1, SW2)
Sensor and Actuator Control Connector(9 )
LED BAR display, Microphone input, Speaker output
DC 12V Power Supply Connector
(3) Base devices
PIR Motion Sensor Block time: 2.5S (default)
Delay time: 5 S (default)
Sentry Angle: < 110 degree
Sentry Distance: 3m (default) - max 7m
Lens Size : Diameter : 23mm (Default)
ILUM Sensor CDS 10mm Light Sensor
TEMP/HUMI Sensor Temperature: -40~80℃ Range, Digtal input I/F
(DHT11) Humidity: 0~100% Range, Digtal input I/F
Smoke Sensor (Gas) Interface type: Analog, Wide detecting scope
Wide detecting scope
DC BLOWER FAN 51 X 52mm DC BLOWER FAN
LED 10mm Round type LED(4EA)
KEYPAD 3-COLUMN, 4-ROW Push Button Switch
BUZZER Interface type: Analog
wide detecting scope
TEXT LCD 16 X 2 Text LCD

(4) External Device
Display 7" IPS Screen LCD
1024 x 600 hardware resolution
Capacitive Touch Control
WebCam HD Web Cam C310
Max resolution: 720p / 30fps
Angle: 60 degree
Microphone CMP-G7 USB Microphone
Direction: Omni directional
Sensitivity : 20 ± 4dB

2) Software specs
(1) Linux Kernel Version : Linux RaspberryPi 4.19.97-v7l+
(2) Rasbian OS Version : Rasbian GNU/Linux 10 (buster)
(3) Google Assistant: Google Assistant library 1.0.1, Google Assistant grpc 0.2.1
(4) OpenCV : opencv-python 3.4.4, opencv-contrib-python 3.4.3
(5) TensorFlow : TensorFlow 1.14.0
(6) ALL Device Demo Program: mCube-AI control
(7) Provides video recognition demo program based on Google Vision and OpenCV
① Google Vision-based image analysis and prediction
* GV Face Detection: Recognizing human faces in images
* GV Facial Expression Detection: Recognition of emotional states on human faces in images
* GV Image Properties Detection: Detects the part's color of the object shown in the image
* GV Label Detection: Recognize each element in the image
* GV Landmark Detection: Recognize the name and location of the building in the image
* GV Safe-Search Detection: Recognize the safety status of the person in the image
* GV Text Detection : Text detection in images
② Face recognition through OpenCV-based image processing and face-recognition library
* OpenCV Control Device by Face Recognition: LED blinks when recognizing an image learned through the camera
* OpenCV Face Detection: Detecting human faces through the camera
* OpenCV Face detection in the picture: Recognition of a given person in the image
* OpenCV Face Recognition: Recognition of the person determined through the camera
* OpenCV Face Recognition from WebCam: Recognition of a number of learned characters through the camera
③OpenCV-based image processing and numeric recognition through Tensorflow
* OpenCV Number Recognition: Learn Mnist numeric data through Tensorflow, recognize the numbers in the handwritten image file, and output the numbers to the speaker.
(8) Voice Recognition Device Control based on Google Assistant
① Device Test Application Source based on voice recognition: Individual actuator control and sensing through voice recognition
* led.py, keypad.py, fan.py, textlcd.py, buzzer.py, cds.py, gas.py, dht11.py, motion.py,
② Application Programming Source based on voice recognition: Various actuator control and sensing through voice recognition
* fire_detect.py, air_condition.py, light_control.py, security.py
(9) IoT-based Application Programming Source: LED control and sensor value sensing through Android App
① adc.py, _led.py, deviceserver.py
(10) Tensorflow machine learning theory and practice program is provided
① Machine learning basics
② Linear regression
③ Logistic regression
④ Softmax
⑤ CNN
(11) Tensorflow-based Machine learning Programming Source
: A practical program on the theory of linear regression and logistic regression, methods of implementing them, and algorithms
① linear_regression.py, cost_graph.py, gradient_descent.py, gradient_descent_tf_optimizer.py,
multi_variable_linear_regression.py, multi_variable_matmul_linear_regression.py, input_fiile_data.py,
logistic_regression.py
(12) Tensorflow-based Deep learning Programming Source
: Deep learning theory using Mnist data, a practical program on how to implement it, and algorithms
① softmax.py, mnist_softmax.py, mnist_nn.py, mnist_nn_xavier.py, mnist_nn_dropout.py, mnist_cnn.py
(13) Tensorflow-based image recognition deep learning Programming Source
: Recognize flowers through lower image data deep learning
① label_image.py, retrain.py