Edge AI

Neural Network / Classification

Good morning, I am working on this Firmware https://github.com/stm32-hotspot/STM32WB-BLE-AI-MotionSense.I created the Neural Network but I don't see the Classification on The Monitor.Can you help me to solve the problem ?Thank you.Best regardsLuigi B...

Keras 2.12.0 support

I want to know when will stm cube ai support keras 2.12.0 because for now it only support up till 2.11.0. I can't seem to downgrade my keras version as it require to downgrade more dependencies.Furthermore I converted my lstm keras (h5) model to tfli...

INTERNAL ERROR: tuple index out of range when using Conv1d and ConvTranspose1d

Hi, I have two onnx models. One of them is U-Net like architecture and another is just Conv1d + Linear. I have onnx file and try to analyze the models.When I analyze U-Net like model, the analyzer show "INTERNAL ERROR: tuple index out of range" error...

Error analyzing a model with custom layers

I am using the XCUBE-AI software to deploy an SNN (Spiking Neural Network) on my STM32H747I-DISCO board. The SNN is in the form of a .h5 file, and was converted from a CNN using the snn toolbox. The SNN ofcourse, has a lot of custom layers. I want to...

E200(ValidationError): stm32: Unable to bind the STM AI runtime with "network" c-model: []  connection to "serial:115200"/"115200" run-time fails  E801(HwIOError): Invalid firmware - C

When I try to Validate model on target using CubeIDE tool I am facing this error.

running on target validation on STM32Cube IDE and XCube AI

I have a saved tflite model , which I want to deploy on my STM32H747I-DISCO board. For now, I just want to run validation on target using one custom input. I have the input as a numpy array (saved as a .npy file, it's a 4D array). Can someone please ...

When I use stm32cubeMX to generate a project, using keil to compile it fails?(It is about x-cube-AI.)

The API explaination of x-cube-Ai

Now I am working on x-cube-ai on STM32F407,but I find little information about the APIs,please help me. Thanks!

Access non contiguous RAM slots in Arduino Nicla Vision

I'm working on a TinyML project with an Arduino Nicla Vision (STM32H747AII6) and I need to store a model with size 608kB in RAM. In the Nicla Vision datasheet, it is stated that 864kB of SRAM are available, but it is segmented in three non contiguous...