Deployment on ESP32 from python


Project ID:

Context/Use case:

Hi I have this model in python that I need to deploy on ESP32 but I am not sure how to do it. I am really stuck and need some guidance. Here is a link the code
Please let me know if you can help out with this. It could literally save my capstone project.

Hi @gsawalha,

The notebook you linked is empty.

To work with Edge Impulse, you will need to create a project, put your data into that project, and train a model using the Edge Impulse Studio. See here for information on how to do that: Getting Started - Edge Impulse Documentation

Once you have a trained model in Edge Impulse, you can deploy it as C++ code or as a pre-compiled firmware for your ESP32. See here for more information: Espressif ESP-EYE (ESP32) - Edge Impulse Documentation

Apologies for the mistake with the link here is the new one kindly take a look Sdp2022 – Google Drive

Hi @gsawalha,

The easiest thing to do right now would be to import your data into an Edge Impulse project, reconstruct the NN model (it looks like you’re using a simple DNN) either with the GUI or in expert mode, and then re-train inside Edge Impulse. From there, you can deploy C++ code as a library to use in your ESP32 project.

You can also try the “bring your own model” tool (Bring your own model (BYOM) - Edge Impulse Documentation), but it’s quite involved at the moment (as you need to create your own Docker image). As a result, I recommend the first method I mentioned.

1 Like