Simple Voice Recognition System Using Arduino and Raspberry Pi
The goal of this project was to utilize both an Arduino and Raspberry Pi in a Voice recognition system.
It only recognizes the words "red", "green", and "blue", then lights up corresponding LEDs.
Life would have been easier to just use the Raspberry Pi with a webcam mic and voice recognition software like Pocket Sphinx, but again the challenge was to use both devices.
The task was divided up by using the Arduino for audio sample capture and audio thresholding.
Raspberry Pi then handled the majority of the voice recognition.
The concept used here was to split the audio into 4 phonemes, transform each into the frequency domain, use the content as features into a neural network, then finally light up the correct LED.