How to extract audio objects with deep learning – without explicitly learning to extract those? In our ICASSP paper we propose multichannel-based learning, a technique closely related to self-supervised learning, differentiable digital signal processing, and universal sound separation.
arXiv: https://arxiv.org/pdf/2102.06142.pdf
Our ICASSP paper studying permutation ambiguity on speaker-independent source separation models is now accessible on arXiv:
https://arxiv.org/pdf/2102.04945
These are the papers we will be presenting at ICASSP 2021:
Infinite thanks to all my collaborators for the amazing work 🙂
Yesterday, I presented our work at the London Audio & Music AI Meetup and in a couple of weeks I’ll be also presenting at the Perth Machine Learning Group Meetup. The slides I’m using for those presentations and the recording of the video are now available online. Hopefully, they provide an additional perspective to our paper!
Upsamplers are a key element for developing computationally efficient and high-fidelity neural audio synthesizers. Given their importance, together with the fact that the audio literature only provides sparse and unorganized insights, our work is aimed at advancing and consolidating our current understanding of neural upsamplers.
