input-speech is an instrument that uses its microphone for recognizing voice commands, to train a neural network built with the Google TensorFlow library. This allows you to control different multimedia outputs.
The instrument can be programmed with a computer and then run as a standalone instrument, or you can use its serial port to communicate with your computer for debugging or interfacing with other software.
Follow the instructions on installation.md.
Follow the instructions on wiring.md.
hello_tiny_world: this is an example that cycles through all the LEDs and their colors.
Record .wav files with this format:
- 1 audio channel (mono)
- 16 kHZ sample rate
- 16 bits per sample: 16 bits
Open your terminal and go to the instruments folder.
cd instrumentsActivate the virtual environment
source env/bin/activateRun Jupyter lab
jupyter-labOpen the window on your web browser, and navigate to the gesture notebook.
Open the Arduino sketch