Edge AI

..

..

..

..

..

..

..

..

On May 12th an interesting Webinar was presented on the smart use of the Machine Learning Core embedded in the X-family IMU sensors, such as the LSM6DSOX mounted on the SensorTile.box board.The main example proposed during the webinar is related to the use of the SensorTile.box together with ST tools (Unico, Unicleo and 

Hi everyone,I am trying to implement an image recognition application on an STM32L475. I succeded implementing a Keras model with Cube AI and wanted to reduce the memory footprint with a quantization model. I used the following code to quantize the keras model according to a question I found on this forum.# Model_s is a trained model in the same python script converter = tf.lite.TFLiteConverter.from_keras_model(model_s) converter.target_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] converter.inference_input_type = tf.int8 converter.inference_output_type = tf.int8 quant_model = converter.convert() # Saving the model open("Models/converted_model.tflite", "wb").write(quant_model)When I try to analyze the model Cube AI reports the following error:Analyzing model Neural Network Tools for STM32 v1.2.0 (AI tools v5.0.0) -- Importing model INTERNAL ERROR:  Creating 

Hello everyone, I would like to ask you about a handwritten number recognition project I have been working on recently. How to input the collected 28*28 image data into the neural network and use keil programming to verify it on the stm32f407 board? Or are there any relevant articles to recommend? Thank you very much!

..

HiI am using voltage and current sensors for state estimation of storage system. I want to do this state estimation through NN. Is it possible to use this STM32Cube AI for train and test the data. If it is possible,please kindly provide the help with timely manner.

..

Hello,is it possible to import weights at runtime? I plan to convert a NN with Cube AI and deploy it to the controller, and later on receive weights via BT and overwrite the previous weights. Would that be possible or can the weights only be set while converting the network with CubeAI?I couldn't find any information on this so far...

This forum seems to be new, we all have problems, everyone looks and nobody brings solutions. The official support only gives code to verify the neural network, but I do not want to verify, I want to predict, I want to use the neural network in the microcontroller, I have already passed the verification stage. Some help please with the functions and initialization to be able to use the neural network. regards

I have encountered error while executing validation on my first Keras model using XCUBE.AI extension. The error message below,----------------------------------------------------------------------------------------------------------------------------------Analyzing model Neural Network Tools for STM32 v1.2.0 (AI tools v5.0.0) TOOL ERROR: Traceback (most recent call last):   File "C:\Users\moham\STM32Cube\Repository\Packs\STMicroelectronics\X-CUBE-AI\5.0.0\Utilities\windows\tensorflow_core\python\pywrap_tensorflow.py", line 58, in <module>     from tensorflow.python.pywrap_tensorflow_internal import *   File "C:\Users\moham\STM32Cube\Repository\Packs\STMicroelectronics\X-CUBE-AI\5.0.0\Utilities\windows\tensorflow_core\python\py

I have OV7670 camera and Nucleo-L496ZG development board, Is it possible to run AI model using them ?? I am trying to achieve object detection using these two. If yes please let me know how (Probably an example code would be great). Thank you in advance

Hello,I successfully used a post-training quantized TFLite Model in X-CUBE-AI. Nonetheless, I can't make it work with a quantization aware training model, because I get the following error when I try to analyze or generate code : " TOOL ERROR: Unknown layer type FAKE_QUANT, stopping ".Is X-CUBE-AI compatible with quantization aware training model ? If so, how am I supposed to export the model from TFLite so that it doesn't contain FAKE_QUANT layers ?Sincerely

Hi all,I recently got a STM32H747 DISCO kit, however, this is my first time using stm MCUs, and the programming and debugging on here is new to me. My main purpose for this kit is to use it to implement a person detection NN. Could someone show me a good starter example for this kit? thank you

How do you see an STM32 Prediction output after implementing a pre-trained model with MXCUBE and validating it on desktop and on target?https://www.youtube.com/watch?v=grgNXdkmzzQ&t=So i accomplished all the validation concerning the HAR CNN Keras Model , how can I see an obvious output of the prediction of the system ? More explicitly: if i enter a custom data of a person "Walking" how can I see that the system has analyzed the data and predicted that the person is indeed walking...

Hello, I am trying using FP-AI Sensing on STEVAL-STLKT01V1 sensortile board. I enabled UART as per the instruction by enabling SENSING1_USE_PRINTF in SENSING1_config.h and SENSING1_USE_PWR_MGNT to 0. I also did 10ms delay for initialization process. But I donot recieve anything in the console after everthing followed as per the instruction. Should there be any changes to be made?

HiI am trying to run artifical intelligence project with STM32F4. Firstly, I ran the on project by following the steps from this site ( https://www.dlology.com/blog/how-to-run-deep-learning-model-on-microcontroller-with-cmsis-nn/ ), Secondly, I necessary to run a one-dimensional project so The project in the (https://github.com/mscuttari/embedded-sound-classifier) site I tried to run the project with stm32f4 but I can not run. How I can run this project in the first project format? ThanksRegards

I am trying to create an AUDİO CLASSIFICATION(https://github.com/mscuttari/embedded-sound-classifier) project in STM32F407 DİSCOVERY. I am trying to follow the steps in the https://www.youtube.com/watch?v=grgNXdkmzzQ&feature=youtu.be link and I get this error. How do I fix this error?Thank you

Hello, I am trying to use the MLC on the LSM6DSOX with the 6D Position Recognition Sample.The Interrupts on INT1 are triggered as expected, but when I read the MLC0_SRC (0x70U) it always returns 0. Moreover, I tried different samples - same problem.I was able to read other registers, for example the TAP_SRC where the data is properly returned. (Double Tap/Single Tap)So, is there a difference in reading the MLC0_SRC and TAP_SRC except the address?Thanks in advance, Manuel