 |
|
Related Communities:
|
 |
|
 |
Music Genre Classification Based on Signal Processing
| Author(s): | Evstifeev S., Shanin I. |
| Published: | Selected Papers of the XX International Conference on Data Analytics and Management in Data Intensive Domains (DAMDID/RCDL 2018). CEUR Workshop Proceedings, Vol. 2277. P.157-161. 2018. |
| Abstract: | |
| Music genre is a description, that allows to categorize music compositions into broader
categories with similar characteristics. With the development of streaming platforms (iTunes music,
SoundCloud, Spotify), the automatic classification of music is becoming increasingly important as a way to
intelligently search in a large number of music files, and also as a support in building recommendation
systems. In this paper, this approach is based on the extraction of information from a signal (timbre, rhythm,
melody, pitch), as well as the construction of high-level features with subsequent classification by methods
of machine learning, in particular, the gradient boosting trees and neural networks are considered. The
GTZAN dataset is used to evaluate the performance of the algorithms with the best result of 78% precision.
Comparison of algorithm with third-party systems is considered. To test algorithms on real data, a website
has been developed that allows to automatically classify users’ music. |
| Download: |
[ http://ceur-ws.org/Vol-2277/paper28.pdf ]
|
|
|
