Hi @tom239955_st ,The `STM32H747I-DISCO_Webcam_V300.bin` was built using IAR Embedded Workbench from the FP-AI-VISION1 v3.0.0 sources. I did a quick test using STM32CubeIDE v1.6.1 and did not see any issues.The STM32H747-Disco LCD screen is not bei...
In your CubeMX project, make sure to enable the X-CUBE-AI Application -> Validation component (UM2526 rev7 Figure 21, page 16). You generated code should contain a file named aiValidation_TFLM.c and look for the aiPbCmdNNRun() function.
Perhaps you could try running a X-CUBE-AI Application ->Validation with the tflite model you created. The generated application will show you how to feed the input tensors and read the output tensors over UART.
In order to use the TFLm runtime, the Keras .h5 needs to be converted in to a .tflite file. Details on how to do this can be found in the X-CUBE-AI html documentation: X-CUBE-AI_install_dir/X-CUBE-AI/6.0.0/Documentation/quantization.html#ref_tf_supp...
