Build and run an image classification NN model on an STM32L4 Discovery board
Introduction
Prepare environment
Build an image classification NN model trained with the CIFAR-10 dataset
Deploy the image classification NN model on STM32
Run the image classification NN model on STM32
Review
Next Steps

  Log an issue

  Fork and edit

  Discuss on Discord

In this section, you will deploy the model directly on the STM32 board.

Setup STM32CubeMX project

STM32CubeMX is a graphical tool for configuring STM32 microcontrollers.
STM32Cube.AI is an extension of STM32CubeMX to import ML models.

Troubleshooting
  • The testing tool (see later) has been written with X-CUBE-AI 7.0.0. Please select this version when installing. Later versions may not connect successfully.

  1. Download the Windows installer for STM32CubeMX from the STM32CubeMX product page .

  2. Run the Windows installer and install in desired location of your Windows Machine.

  3. Open STM32CubeMX IDE.

  4. Click Access to Board Selector. Find your B-L475E-IOT01A board and click Start Project.

  5. Go to Project Manager. Enter a project name and select the project location where the project will be saved.

  6. Under Pinout & Configuration, expand Pinout menu, and click Clear pinouts.

  7. Now install Cube AI and enable it for your project.

  8. Expand Software Packs menu, and click Select Components.

  9. Locate X-CUBE-AI, and click its Install button.

  10. Expand the menu and enable X-CUBE-AI > Core. For device application, choose Validation. If the status shows a yellow warning, click Resolve to install any necessary dependencies.

  11. Click OK when done.

Deploy the model with STM32CubeMX

  1. Open STM32CubeMX.

  2. Click Access to Board Selector.

  3. Find the B-L475E-IOT01A2 board and click Start Project. Accept any defaults.

  4. Go to Project Manager. Set the project name and save location.

  5. Set Toolchain/IDE as STM32CubeIDE

Image Alt Text:output3

  1. Go to Pinout & Configuration and clear pinouts from the Pinout menu.

Image Alt Text:output4

  1. In Software Packs menu, click Select Components. Enable X-CUBE-AI. For device application, choose Validation. Click OK to save.

Image Alt Text:output5

  1. Navigate to X-CUBE-AI configuration.

  2. Click Add network. Browse to the model provided (Data/models/cifar10_model.h5) or the model you trained by yourself (Data/models/own_cifar10_model.h5).

  3. Click Analyze to check the model.

  4. Generate the validation code for the model by clicking Generate Code.

Image Alt Text:output6

  1. Open STM32CubeIDE.

  2. Navigate to File > New > Project > ST > STM32 Project from an Existing STM32CubeMX Configuration File.

  3. Go to the project folder and open the .ioc file.

  4. Ensure that the board is connected to your computer. If it is correctly connected, build and flash the code by clicking Run As.

Image Alt Text:output7

  1. If you get an ‘undefined reference’ error, go to Core/Src/main.c. Remove static from the declaration of the MX_USART1_UART_Init() function and also from its definition. Try Run As again.

Image Alt Text:output8

With the model now deployed on the STM32 board, you are ready to test it.

Back
Next