This program is a tool that provides a transcription of notes given an audio recording, created by the user. This program uses artificial intelligence to recognize the notes, and can be useful to for a variety of reasons:
- Checking the tune of an instrument
- Confirming whether you're singing in tune
- General transcription of a melody
Python version == 3.10.14
Install all dependencies in requirements.txt. Note that requirements.txt is meant for macs using silicon chips. However, on any other platform, ensure that tensorflow version >= 2.0.0 is installed in the ran environment, along with the depenencies listed in requirements.txt.
If ran on a mac, ensure tensorflow is installed correctly. More information can be found here for intel chips and here for silicon chips
To run the program, run main.py. The program will be interacted via the console.
The program works best with slow, single melodies with no background noise or tracks. As a good rule of thumb, if a melody can be hummed clearly, the program will run well.