• Install Arduino IDE

• Open Arduino IDE after installation

• Enter Preferences page

• Add one or more JSON links

https://espressif.github.io/arduino-esp32/package_esp32_index.json

• Change the project folder to C:\Users\Waveshare\AppData\Local\Arduino15\packages (Waveshare is the computer username)

• Enter development board manager, search for esp32, select Version 3.0.2 from esp32 by Espressif Systems below, and click INSTALL (if unable to install properly, you can try using a phone hotspot)

• Restart Arduino IDE after installatio

• Please note whether the library has been installed before. If it has been installed, please archive the original library to prevent errors and failures when running other programs.
• Enter the example folder ESP32_S3_Touch_LCD_1.85\LVGL_Arduino, and double click LVGL_Arduino.ino to open the example.
• Search for lvgl library and install it (this operation is to set up a library folder in Arduino environment. If other libraries have already been installed before and following folder exists, this step is not necessary).

• Enter the path of library file c:\Users\Waveshare\AppData\Local\Arduino15\packages\libraries (Waveshare is the computer username). If you are unsure which folder it is in, please go to Preferences to check.

• If you don't know the path, you can check it as below:

• Delete the downloaded lvgl library (if not installed earlier, ignore this step)

• Install the two libraries provided and create a new Lib folder in the example path

• Download 2 library files (https://files.waveshare.com/wiki/common/Lvgl.zip lvgl library] and audioI2S library) at the bottom of this page, uncompress and save it in the newly created Lib path
• Copy the library provided to c:\Users\Waveshare\AppData\Local\Arduino15\packages\libraries (Waveshare is the computer username)

• Go on and install NTPClient library

• Install SensorLib library

• Complete library installation

   English recognition environment setting

• The initial state of the environment defaults to English recognition environment. If it is not converted to Chinese, the following modifications will not be made
• If it has been modified to Chinese recognition before, download the provided program file Use_the_English_model_program
• Enter the path C:\Users\Waveshare\AppData\Local\Arduino15\packages\esp32\hardware\esp32\3.0.2\libraries\ESP_SR\src (Waveshare is the computer user name)
• Use the downloaded file to replace esp32-hal-sr.c in this path

• Download English model English_model_-_wake-up_words (hi_esp)
• Enter the path C:\Users\Waveshare\AppData\Local\Arduino15\packages\esp32\tools\esp32-arduino-libs\idf-release_v5.1-bd2b9390ef\esp32s3\esp_sr (Waveshare is the computer user name)
• Use the downloaded file to replace srmodels.bin in this path
• Modify the recognition instructions in the program to English instructions (phonemes)

   Generate English phonemes

• Download Generate_English_command file
• Click Win + R to enter the command window, and use "cd path" to enter the downloaded folder
• Install en_g2p

pip install en_g2p

• Execute miltinet_g2p.py file

python miltinet_g2p.py

• Execute gen_sr_commands.py to generate phoneme instructions

python gen_sr_commands.py "English to be converted into phonemes"
Example: python gen_sr_commands.py "hello word"

   Chinese recognition environment setting

• The initial state of the environment defaults to English recognition environment. If it has not been modified to Chinese recognition before, download the provided program file Use_the_Chinese_model_program
• Enter the path C:\Users\Waveshare\AppData\Local\Arduino15\packages\esp32\hardware\esp32\3.0.2\libraries\ESP_SR\src (Waveshare is the computer user name)
• Use the downloaded file to replace esp32-hal-sr.c in this path

• Download Chinese model Chinese_model_-_wake-up_words (hi_乐鑫)
• Enter the path C:\Users\Waveshare\AppData\Local\Arduino15\packages\esp32\tools\esp32-arduino-libs\idf-release_v5.1-bd2b9390ef\esp32s3\esp_sr (Waveshare is the computer user name)
• Use the downloaded file to replace srmodels.bin in this path
• Modify the recognition instructions in the program to Chinese instructions (pinyin)

   Generate Chinese pinyin

• Download Generate_Chinese_command file
• Click Win + R to enter the command window, and use "cd path" to enter the downloaded folder
• Execute multinet_pinyin.py to generate Chinese pinyin

python multinet_pinyin.py "Chinese to be converted to pinyin"
Example: python multinet_pinyin.py "你好新中国"