To install required dependencies please run :
pip install -r requirements.txt- The Database consists of 223 songs, recognition is based on songs spectrograms or the spectral features :
( spectral centroids & spectral rolloff ), hashing the data is done with a Perceptual Hashing algorithm. - Two audio files can be loaded into the application, you can mix the 2 files as a testing mechanism.
- The results are populted into the table and sorted according to their similarity index
