Final project for the CEN598 course of Arizona State University. We showcase the strength of SFL in Arduino Nano 33 BLE Sense.
Report is arxived here: Split Federated Learning on Micro-controllers: A Keyword Spotting Showcase
- Extend the server-side model to more complex architecture
- Extend the serial communication via bluetooth
''' pip install pyserial numpy matplotlib '''
- Open the project with PlatformIO and flash the firmware to all the boards
- Run the sfl_server.py using Python3 (This is a simulated training process). The default port should be /dev/ttyACM0 or /dev/ttyACM1
- Checkout "fl" branch of this repo
- Run the fl_server.py using Python3 (This is a simulated training process). The default port should be /dev/ttyACM0 or /dev/ttyACM1
- Jingtao Li
- Runcong Kuang
This project is originated from Federated Learning with Arduino Nano 33 BLE Sense, which is authored by:
- Marc Monfort
- Nil Llisterri