EdgeAI-AMB82mini

EdgeAI AMB82mini

教材網址：https://github.com/rkuo2000/EdgeAI-AMB82mini

LLM服務器程式範例： [AmebaPro2_server/*.py](https://github.com/rkuo2000/EdgeAI-AMB82mini/blob/main/AmebaPro2_server)
Arduino 程式範例： [Arduino/AMB82-mini/](https://github.com/rkuo2000/EdgeAI-AMB82mini/blob/main/Arduino/AMB82-mini)

自強基金會 WiFi

SSID: TCFSTWIFI.ALL
Pass: 035623116

1. AI 介紹

2. 開發板介紹

RTL8735B晶片簡介

32-bit Arm v8M, up to 500MHz, 768KB ROM, 512KB RAM, 16MB Flash (MCM embedded DDR2/DDR3L up to 128MB)
802.11 a/b/g/n WiFi 2.4GHz/5GHz, BLE 5.1, NN Engine 0.4 TOPS, Crypto Engine, Audio Codec, …

3. IDE使用介紹

偏好設定 (Preferences)

Hub8735 ultra
https://raw.githubusercontent.com/ideashatch/HUB-8735/main/amebapro2_arduino/Arduino_package/ideasHatch.json

AMB82-mini
main https://github.com/Ameba-AIoT/ameba-arduino-pro2/raw/main/Arduino_package/package_realtek_amebapro2_index.json
dev https://github.com/Ameba-AIoT/ameba-arduino-pro2/raw/dev/Arduino_package/package_realtek_amebapro2_early_index.json

選定開發板 AMB82-MINI

Tools > Board Manager > AMB82 package > 4.0.9

Serial-monitor = 115200 baud

Getting Started

首先將AMB82-mini板子用MicroUSB線連接至電腦的USB port
確認UART com port (Ubuntu OS需 sudo chown usrname /dev/ttyUSB0)

燒錄程式碼：

按下UART_DOWNLOAD按鈕, 再按下RESET按鈕, 放開RESET按鈕, 再放開UART_DOWNLOAD按鈕,板子就進入燒錄模式.
然後于Arduino IDE上按下燒錄按鍵Upload

Realtek AmebaPro2 hardware libraries

C:\Users\user\AppData\Local\Arduino15\packages\realtek\hardware\AmebaPro2\4.0.9-build20250805\libraries

C:\Users\user\AppData\Local\Arduino15\packages\realtek\hardware\AmebaPro2\4.0.9-build20250805\libraries\NeuralNetwork\src
GenAI.h
GenAI.cpp

Arduino examples 範例練習

Examples> 01.Basics > Blink

Examples> 02.Digitial > GPIO > Button

程式碼修改：
const int buttonPin = 1; // the number of the pushbutton pin
const int ledPin = LED_BUILTIN; // the number of the LED pin

Examples> 01.Basic > AnalogReadSerial

程式碼修改：Serial.begin(115200);

AMB82-Mini 程式範例

瀏覽器打開 EdgeAI-AMB82mini, 點[Code]並選 [Download ZIP]
解壓縮.zip, 並將 Arduino/AMB82-mini 複製到文件/Arduino下, 成為 文件/Arduino/AMB82-mini

4. Examples 範例練習

MCU Interfaces

WiFi

Examples> WiFi > SimpleTCPServer
WiFi - Simple TCP Server

Examples> WiFi > SimpleHttpWeb > ReceiveData
WiFi - Simple Http Server to Receive Data

Examples> WiFi > SimpleHttpWeb > ControlLED
WiFi - Simple Http Server to Control LED

Sketchbook> AMB82-mini > WebServer_ControlLED
Sketchbook> WebServer_ControlLED

BLE

Exmples> AmebaBLE > BLEV7RC_CAR_VIDEO

BLE V7RC

V7RC APP 介紹
安裝手機App V7RC

MQTT

MQTT is an OASIS standard messaging protocol for the Internet of Things (IoT)
How MQTT Works -Beginners Guide

Examples> AmebaMQTTClient > MQTT_basic
MQTT - Set up MQTT Client to Communicate with Broker

pip install paho.mqtt

publish messages to AMB82-mini

import paho.mqtt.publish as publish
host = "test.mosquitto.org"
publish.single("ntou/edgeai/robot1", "go to the kitchen", hostname=host)

subsribe messages from AMB82-mini

import paho.mqtt.subscribe as subscribe
host = "test.mosquitto.org"https://github.com/rkuo2000/EdgeAI-AMB82mini
msg = subscribe.simple("ntou/edgeai/robot1", hostname=host)
print("%s %s" % (msg.topic, msg.payload.decode("utf-8")))

Gemini MQTT App

Google Gemini +Canvas

Prompt: make an html to input MQTT topic and text to publish through Paho-MQTT test.mosquitto.org

Paho-MQTT Publisher

5. 感測器與週邊裝置

紅外線測距模組

VL53L0X v2 , VL53L1X v2

Datasheet: VL53L0X - Time-of-Flight ranging sensor

Sketchbook> AMB82-mini > IR_VL53L0X

慣性感測

慣性感測元件

MPU6050

Sketchbook > AMB82-mini > MPU6050-DMP6v12

PWM

PWM - Servo Control

Examples> AmebaAnalog > PWM_ServoControl

myservo.attach(8);

myservo.write(pos);

H-bridge 全橋式馬達驅動

TB6612

DRV8833

ILI9341 TFT-LCD

SPI - LCD Screen ILI9341 TFT

Interface signal names:

MOSI: Standard SPI Pin
MISO: Standard SPI Pin
SLK: Standard SPI Pin
CS: Standard SPI Pin
RESET: Used to reboot LCD.
D/C: Data/Command. When it is at LOW, the signal transmitted are commands, otherwise the data transmitted are data.
LED (or BL): Adapt the screen backlight. Can be controlled by PWM or connected to VCC for 100% backlight.
VCC: Connected to 3V or 5V, depends on its spec.
GND: Connected to GND.

AMB82 MINI and QVGA TFT LCD Wiring Diagram:

Sketchbook> AMB82-mini > HTTP_Post_ImageText_TFTLCD

Exmples> AmebaSPI > Camera_2_lcd
Camera output , then Jpeg Decoder to TFT-LCD

compilation error: need to modify Libraries/TJpg_Decoder/src/User_Config.h
//#define TJPGD_LOAD_SD_LIBRARY

Exmples > AmebaSPI > Camera_2_Lcd_JPEGDEC
Camera output, saved to SDcard, then Jpeg Decoder to read to TFT-LCD

Exmples > AmebaSPI > LCD_Screen_ILI9341_TFT
LCD Draw Tests

6. 影像串流範例練習

影像串流

Examples> AmebaMultimedia > StreamRTSP > VideoOnly
Sketchbook> RTSP_VideoOnly

Amebapro2 AMB82-mini Arduino Example Guides

動作偵測 (motion detection)

Examples> AmebaMultimedia > MotionDetection > LoopPostProcessing

修改ssid, passwd, 後燒錄到AMB82-mini,
按reset後程式即開始運行, 用serial-monitor 查看顯示串流網址
啟動手機或電腦上之VLC player, 設定RTSP串流網址

Motion Detection Google Line Notify

Examples> AmebaMultimedia > MotionDetection > MotionDetectGoogleLineNotify

7. 音頻應用

Audio & Mic

音頻環回測試

Examples> AmebaMultimedia > Audio >LoopbackTest

MP3 播放

AMB82-mini + PAM8403 + 4ohm 3W speaker

Sketchbook> AMB82-mini > SDCardPlayMP

.mp3 files stored under mp3 directory

Sketchbook> AMB82-mini > SDCardPlayMP_All

音頻串流範例

Examples> AmebaMultimedia > Audio > RTSPAudioStream

RTSP Audio Stream

MP4錄音範例

Examples> AmebaMultimedia > RecordMP4 > AudioOnly

Multimedia - MP4 Recording

音頻分類範例

Examples> AmebaNN > AudioClassification

YAMNet

8. 人臉辨識與識別

人臉辨識與識別介紹

人臉檢測範例

Examples> AmebaNN > RTSPFaceDetection

人臉識別範例

Examples> AmebaNN > RTSPFaceRecognition

Serial_monitor: REG=RKUO

Enter the command REG=Name to give the targeted face a name.
Enter the command DEL=Name to delete a certain registered face. For example, DEL=SAM
Enter the command BACKUP to save a copy of registered faces to flash.
If a backup exists, enter the command RESTORE to load registered faces from flash.
Enter the command RESET to forget all previously registered faces.

9. 影像分類 (Image Classification)

影像分類範例

RTSP_GarbageClassification.ino

Garbage模型訓練與檔案轉換

模型訓練： kaggle.com/rkuo2000/garbage-cnn

required in kaggle for AmebaPro2 1) pip install tensorflow==2.14.1 2) model.save(‘garbage_cnn.h5’, include_optimizer=False)

模型轉換：AMB82 AI Model Conversion

Download garbage_cnn.h5 from kaggle.com/rkuo2000/garbage-cnn Output
Compress garbage_cnn.h5 to garbage_cnn.zip
Go to Amebapro2 AI convert model, fill up your E-mail
Upload garbage_cnn.zip
Upload one (.jpg) test picture (EX. glass100.jpg from Garbage dataset)
Email will be sent to you for the link of network_binary.nb

程式範例：RTSP_GarbageClassification.ino

click the recieved Email link to download network_binary.nb
create NN_MDL folder in SDcard, save network_binary.nb under NN_MDL folder, and rename it to imgclassification.nb
plugin SDcard back to AMB82-MINI
modify Sketch RTSP_GarbageClassification.ino 1) modify SSID and PASSWD 2) modify imgclass.modelSelect (change DEFAULT_IMGCLASS to CUSTOMIZED_IMGCLASS)
burn code into board AMB82-MINI, and run it with VLC player streaming

10. 物件偵測 (Object Detection)

Public Dataset

Roboflow

物件檢測介紹

Kaggle範例:

Pothole模型訓練與檔案轉換

模型訓練： kaggle.com/rkuo2000/yolov7-pothole

1) repro https://github.com/WongKinYiu/yolov7 2) create pothole.yaml %%writefile data/pothole.yaml

train: ./Datasets/pothole/train/images
val:  ./Datasets/pothole/valid/images
test: ./Datasets/pothole/test/images

# Classes
nc: 1  # number of classes
names: ['pothole']  # class names

3) YOLOv7-Tiny Fixed Resolution Training

!sed -i "s/nc: 80/nc: 1/" cfg/training/yolov7-tiny.yaml
!sed -i "s/IDetect/Detect/" cfg/training/yolov7-tiny.yaml

模型轉換：AMB82 AI Model Conversion

Download best.pt from kaggle.com/rkuo2000/yolov7-pothole
Compress best.pt to best.zip
Go to Amebapro2 AI convert model, fill up your E-mail
Upload best.zip
Upload one (.jpg) test picture (EX. pothole_test.jpg from Pothole dataset)
Email will be sent to you for the link of network_binary.nb

程式範例：RTSP_YOLOv7_Pothole_Detection.ino

click the recieved Email link to download network_binary.nb
create NN_MDL folder in SDcard, save network_binary.nb under NN_MDL folder, and rename it to yolov7_tiny.nb
plugin SDcard back to AMB82-MINI
modify Sketch RTSP_YOLOv7_Pothole_Detection.ino 1) modify SSID and PASSWD 2) modify ObjDet.modelSelect(OBJECT_DETECTION, CUSTOMIZED_YOLOV7TINY, NA_MODEL, NA_MODEL);
burn code into board AMB82-MINI, and run it with VLC player streaming

AMB82 Mini 物件偵測範例

RTSP_YOLOv7_Pothole
RTSP_YOLOv7_Sushi

AMB82 Mini SD卡加載模型範例

RTPS_ObjectDetection_AudioClassification.ino

Online NN Conversion Tool (客製化模型轉換工具)

11. 大型語言模型範例 (LLM)

大型語言模型介紹

語音辨識範例

ffmpeg.exe is needed for Windows to run Whisper!

語音交談範例

ffmpeg.exe is needed for Windows to run Whisper!

Download ffmpeg-master-latest-win64-gpl.zip, extract & put ffmpeg.exe into where you run Whisper server.

AmebaPro2 Whisper LLM_server

AmebaPro2_Whisper_Gemini_server

到 https://aistudio.google.com/app/apikey 取得API_Key 填入GOOGLE_API_KEY, 再執行AmebaPro2_Whisper_Gemini_server.py

HTTP_Post_Audio.ino

修改server IP位址 in RecordMP4_HTTP_Post_Audio.ino server IP位址, then 燒錄到 AMB82-MINI
reset AMB82-MINI 來啟動, 按鍵兩秒後即可錄音詢問 LLM/Gemini

12. 視覺語言模型 (VLM)

視覺語言模型介紹

影像+語音交談範例

13. GenAI

Examples: AmebaNN > MultimediaAI > GenAIVision
Examples: AmebaNN > MultimediaAI > GenAISpeech_Gemini
Examples: AmebaNN > MultimediaAI > TextToSpeech

EdgeAI AMB82mini

1. AI 介紹

2. 開發板介紹

3. IDE使用介紹

Arduino IDE 2.3.6 下載 & 安裝

偏好設定 (Preferences)

選定開發板 AMB82-MINI

Realtek AmebaPro2 hardware libraries

Arduino examples 範例練習

4. Examples 範例練習

WiFi

BLE

Gemini MQTT App

5. 感測器與週邊裝置

紅外線測距模組

慣性感測

PWM

H-bridge 全橋式馬達驅動

ILI9341 TFT-LCD

Interface signal names:

AMB82 MINI and QVGA TFT LCD Wiring Diagram:

6. 影像串流範例練習

影像串流

7. 音頻應用

音頻環回測試

MP3 播放

音頻串流範例

MP4錄音範例

音頻分類範例

8. 人臉辨識與識別

9. 影像分類 (Image Classification)

影像分類範例

Garbage模型訓練與檔案轉換

模型訓練： kaggle.com/rkuo2000/garbage-cnn

模型轉換：AMB82 AI Model Conversion

程式範例：RTSP_GarbageClassification.ino

10. 物件偵測 (Object Detection)

Public Dataset

Pothole模型訓練與檔案轉換

模型訓練： kaggle.com/rkuo2000/yolov7-pothole

模型轉換：AMB82 AI Model Conversion

程式範例：RTSP_YOLOv7_Pothole_Detection.ino

AMB82 Mini 物件偵測範例

AMB82 Mini SD卡加載模型範例

Online NN Conversion Tool (客製化模型轉換工具)

11. 大型語言模型範例 (LLM)

語音辨識範例

ffmpeg.exe is needed for Windows to run Whisper!

語音交談範例

ffmpeg.exe is needed for Windows to run Whisper!

12. 視覺語言模型 (VLM)

影像+語音交談範例

13. GenAI

14. EdgeAI 應用專題實作

邊緣計算微控制器應用專題實作期末報告 (Canva)