Differentiable Wavetable Synthesis
Sound
2022-02-15 v4 Machine Learning
Audio and Speech Processing
Abstract
Differentiable Wavetable Synthesis (DWTS) is a technique for neural audio synthesis which learns a dictionary of one-period waveforms i.e. wavetables, through end-to-end training. We achieve high-fidelity audio synthesis with as little as 10 to 20 wavetables and demonstrate how a data-driven dictionary of waveforms opens up unprecedented one-shot learning paradigms on short audio clips. Notably, we show audio manipulations, such as high quality pitch-shifting, using only a few seconds of input audio. Lastly, we investigate performance gains from using learned wavetables for realtime and interactive audio synthesis.
Cite
@article{arxiv.2111.10003,
title = {Differentiable Wavetable Synthesis},
author = {Siyuan Shan and Lamtharn Hantrakul and Jitong Chen and Matt Avent and David Trevelyan},
journal= {arXiv preprint arXiv:2111.10003},
year = {2022}
}
Comments
Accepted by ICASSP 2022, Demo: https://lamtharnhantrakul.github.io/diffwts.github.io/