WazDat is a simple audio retrieval system based on the algorithms Shazam uses. It can compare a signal with the database in a time invariant, amplitude invariant and noise resistant way. The project is written in python.
A database is created as follows
python main.py -d dbname --train datafolder/subfolder
Files can then be classified using
python main.py -d dbname -f filename
The project can be cloned as follows:
git clone git@github.com:ArnoldSwaggernegger/WazDat.git
Note that the used datasets are stored on the Git Large File Storage (LFS) system.