Does Mbed Team can give a good example for Tinyml in Mbed using Mbed Studio?

My students are studying the Tinyml based on the book " Tinyml: Machine Learning with Tensorflow Lite on Arduino and Ultra-Low-Power Microcontrollers" and some sources from Github.
Actually, he spends almost one Month making the example run on mbed, however, it is failed. He can’t make it work using CLI, very Frustrated.

Ok, Then we try to move to Arduino Nano 33 BLE Sense, only 3 days he can run the example success.

Can Mbed Team give a good example for Tinyml in Mbed using Mbed Studio?
Or we need to give up using Mbed for Tinyml.

Hi @cornetlin, you can do this with Edge Impulse (built by former Arm engineers and by Dan, one of the authors of the book):

  1. Grab data from your devices with https://docs.edgeimpulse.com/docs/cli-data-forwarder - comes with Mbed examples.
  2. Train a model - lots of examples here: https://docs.edgeimpulse.com/docs/getting-started or you can plug in your own Keras code.
  3. Deploy back to the board - https://docs.edgeimpulse.com/docs/running-your-impulse-mbed has examples that run on any Mbed board.
3 Likes

Dear @janjongboom
Thank you so much, forward it to my students.

Hello

can we run 2 tinyML models on arduino nano ble 33 sense? assume using SVM and linear regression