Comparing Audio Features

Thesis Type Bachelor
Thesis Status
Currently running
Student Antonia Kalz
Thesis Supervisor
Research Field

The aim of this work is the comparison of songs by means of audio content descriptors which were determined by different methods. Therefore, e.g., the differences between the 30 second samples and the full version of the song should be examined. Further, it should be investigated which differences may result from using different algorithms and/or frameworks such as e.g., Essentia and Librosa. In addition, the impact of different audio encodings (e.g., WAV vs. MP3) on the resulting features should be investigated.