Skip to content

Latest commit

 

History

History
52 lines (30 loc) · 3.24 KB

README.md

File metadata and controls

52 lines (30 loc) · 3.24 KB

Codacy BadgeCodeQLVercelContinuous IntegrationCypress End2End Tests SpotMyFM

SpotMyFM

A Spotify Library Manager

Check SpotMyFM at spotmyfm.jorgeruizdev.com


¿What is SpotMyFM?

SpotMyFM is a Spotify Library Manager supported by artificial intelligence algorithms.

Please check the Users Manual for a detailed explanation of each of the features.

Some of the features are:

  • Advanced Library Filters. Filter by popularity, moods, LastFM community tags, release date intervals or even genres!
  • Deep Learning Track Analysis. Analyze any track with just a music file! Extract genres, subgenres and even moods😀!
  • Track Recommendatios. SpotMyFM has a small collection of 30k songs to recommend. This recommendations are based on sound features, like Tempo, Beats and their power spectrogram.
  • Playlist Creation. Create or extend an existing playlist with track recommendations or filter results.
  • Library/Playlist Stats. Explore your personal stats like favourite genres, musical taste changes over time or mood and decades distribution.
  • Track/Album/Artist/Playlist details. Navigate through your library! Explore all the tracks of a playlist, albums of an artist or even the tracks of each album with the details view. Read about the history of a band or album, get new recommendations or just check their genres, subgenres or moods!
  • Tag your album. Tag and filter your album with custom strings that identify each album.

And many more!


Datasets

Three datasets were created through the development of this project. Two of them are published at Kaggle

Ludwig Dataset - Genres, Subgenred & Moods - 13k+ Songs (Kaggle)

This is the main dataset that has been used to train all the neural networks

Gtzan Extended Dataset - 3k Songs Dataset (Kagggle)

This dataset has been used to initialize the weights of the main neural network.


Teaser

ezgif-5-abe8f0cb94

Architecture

Poster