Two Edge Impulse model on the same microcontroller (STM32)


I’m trying to deploy two projects on the same microcontroller. First of all, is it possible to deploy them as C++ libraries using the mbedOS IDE?

The idea is to design two neural networks that require the same number of features (thus, the Impulse design, Time Series data, and Processing Block are the same), but with different learning blocks. The microcontroller should acquire a signal, process it with the two different NNs, and produce two different results. It’s not important for the microcontroller to perform both inferences simultaneously.

Is it possible to achieve this? How?

Thank you for any assistance you can provide.


It is currently not possible to run two models on the same device with Edge Impulse. Please see the following: Running multiple (two) Edge impulse model simultaneously in a single device - #4 by aurel

The exception to this is if you deploy a .eim executable to a Linux device, as you can run them in separate processes.