Build error when integrating AI model to project

[LuizF_Ferro]

Hello,

I am developing an AI classification model based on LSM6DSOX sensor data. Before adding the X-CUBE-AI software pack to Cube IDE, I was firstly developing the data acquisition code with the driver from X-CUBE-MEMS1, which I managed to do.

After adding the Cube AI software pack, I uploaded my .h5 file, analyzed and validated the model (both on desktop & target) and I didn't get any errors, so everything was fine.

However, after the generated code was integrated into the project, I started to get several build errors. I didn't change anything in my main.c or in any generated file, so I don't understand why this is happening.

For further information, my MCU is a STM32U585CIU6, and I attached a .txt file with the errors I had.

Could anyone help me, please? Thank you!

[jean-michel.d · Accepted solution]

Hello Luiz,

According the attached file, the build error are related to the specific files requested by the built-in test applications (systemperf and/or validation): bsp_ai.h, aiTestUtility.h, app_x-cube-a.c...

The code/file which is requested to be integrated in a AI application is limited to:

- a set of specialized files (if "network" prefix is used by default) : network.c/h, network_data.c/h. Location

- a static library (NetworkRuntime810_CMx_YYY .a) + associated headers files. Location: Middlewares/ST/AI/[Inc/Lib]

The good starting point is from X-CUBE-AI to generate a Template project (instead aiSystemPerf/aiValidation) for your device and used ARM toolchain to have a simple project with a basic app to use the generated code.

After in the installed X-CUBE-AI (accessible from CubeMX, help-menu) there is a set of documentation/article to explain the usage of the API.

JM