Refer to main.ipynb for plots and figures about our current progress:
- Vocal stems extraction
- Fundamental frequency tracking
- Spectrogram + pitch contour plots
- Data generation with segmented plots
- Data generation with PSOLA autotune algorithm
- Binary classifier with Convolutional Neural Networks
- Training pipeline
- Analysis of results
Setup Instructions:
pip install
conda install -c conda-forge ffmpeg libsndfile
pip install spleeter