Testbed for deep-generative AudioVisuals
- Python 3.8
- Datoviz
- madmom
- carat
- musicsections
- screeninfo
- audioio
- sklearn
Install Datoviz by following https://datoviz.org/tutorials/install/ or
For Linux:
pip install http://dl.datoviz.org/v0.1.0-alpha.1/datoviz-0.1.0a1-cp38-cp38-manylinux_2_24_x86_64.whl
For macOS:
pip install http://dl.datoviz.org/v0.1.0-alpha.1/datoviz-0.1.0a1-cp38-cp38-macosx_10_14_x86_64.whl
For Windows:
pip install http://dl.datoviz.org/v0.1.0-alpha.1/datoviz-0.1.0a1-cp38-cp38-win_amd64.whl
followed by
git clone https://github.com/ahip88/AudioVisual.git
cd AudioVisual
pip install -r requirements.txt
- To run the tutorial:
python "..yourFullPath/AudioVisual/tutorial.py" When_They_Come_For_Me
- For the AudioVisual only:
python "..yourFullPath/AudioVisual/audio_visual_v5.py" sync_test
- To process your own songs, place them in AudioVisual Folder and run:
python "..yourFullPath/AudioVisual/Process.py" File
Do Not include the extension, it will be searched for.
If it's lagging (fps < 300) you're probably being spammed vklite validation warnings. Follow the instructions to build datoviz from Source (updating anything Vulkan related)
ahip88 [email protected]
* Demo/Tutorial