Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/despoisj/DeepAudioClassification
Finding the genre of a song with Deep Learning
https://github.com/despoisj/DeepAudioClassification
artificial-intelligence deep-learning machine-learning music neural-networks
Last synced: about 2 months ago
JSON representation
Finding the genre of a song with Deep Learning
- Host: GitHub
- URL: https://github.com/despoisj/DeepAudioClassification
- Owner: despoisj
- Created: 2016-11-27T12:00:41.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2021-07-26T19:31:50.000Z (about 3 years ago)
- Last Synced: 2024-04-07T02:10:29.220Z (6 months ago)
- Topics: artificial-intelligence, deep-learning, machine-learning, music, neural-networks
- Language: Python
- Homepage: https://medium.com/@juliendespois/finding-the-genre-of-a-song-with-deep-learning-da8f59a61194#.yhemoyql0
- Size: 67.4 KB
- Stars: 1,090
- Watchers: 41
- Forks: 219
- Open Issues: 22
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Deep Audio Classification
A pipeline to build a dataset from your own music library and use it to fill the missing genres
Read the [article on Medium](https://medium.com/@juliendespois/finding-the-genre-of-a-song-with-deep-learning-da8f59a61194#.yhemoyql0)
Required install:
```
eyed3
sox --with-lame
tensorflow
tflearn
```
- Create folder Data/Raw/
- Place your labeled .mp3 files in Data/Raw/
To create the song slices (might be long):
```
python main.py slice
```
To train the classifier (long too):
```
python main.py train
```
To test the classifier (fast):
```
python main.py test
```
- Most editable parameters are in the config.py file, the model can be changed in the model.py file.
- I haven't implemented the pipeline to label new songs with the model, but that can be easily done with the provided functions, and eyed3 for the mp3 manipulation. Here's the full pipeline you would need to use.
