This repository contains an implementation of speech recognition using mel-frequency cepstral coefficients and dynamic time warping on an STM32F4 microcontroller.
You must have the following tools installed on your computer:
- arm-none-eabi-gcc, version 7 or later
- CMake, version 3.9 or later
- make or another build system supported by CMake
- OpenOCD 0.10.0 or later
For Windows users there is a Powershell script scripts/windows_install.ps1
that will install all the above using the Scoop package manager.
Additionally, Python 3 is needed with the following packages to run some code-generation and visualization scripts:
- numpy
- matplotlib
- pyqtgraph
- dtw
- serial
To build:
mkdir build
cd build
cmake ..
cmake --build . --target upload