Marco Comunità (@marcomunita) 's Twitter Profile
Marco Comunità

@marcomunita

PhD researcher in AI & Music @c4dm. Previously @SonyCSLParis @SonyAI_global @AXDImperial @Blackstaramps

ID: 985523285185220610

linkhttps://mcomunita.github.io/ calendar_today15-04-2018 14:20:36

58 Tweet

369 Followers

211 Following

Marco Comunità (@marcomunita) 's Twitter Profile Photo

#AFXResearch update - added: DDSP Guitar Amp: Interpretable Guitar Amplifier Modeling 📜: arxiv.org/abs/2408.11405 🔗: mcomunita.github.io/AFX-Research

#AFXResearch update - added:

DDSP Guitar Amp: Interpretable Guitar Amplifier Modeling

📜: arxiv.org/abs/2408.11405
🔗: mcomunita.github.io/AFX-Research
Marco Comunità (@marcomunita) 's Twitter Profile Photo

#AFXResearch update - added: Hyper Recurrent Neural Network: Condition Mechanisms for Black-box Audio Effect Modeling 📜: arxiv.org/abs/2408.04829 🔗: mcomunita.github.io/AFX-Research ⭐️🙏: github.com/mcomunita/AFX-…

#AFXResearch update - added:

Hyper Recurrent Neural Network: Condition Mechanisms for Black-box Audio Effect Modeling

📜: arxiv.org/abs/2408.04829
🔗: mcomunita.github.io/AFX-Research
⭐️🙏: github.com/mcomunita/AFX-…
Marco Comunità (@marcomunita) 's Twitter Profile Photo

#AFXResearch update - added: Comparative Study of State-based Neural Networks for Virtual Analog Audio Effects Modeling 📜: arxiv.org/abs/2405.04124 🔗: mcomunita.github.io/AFX-Research ⭐️🙏: github.com/mcomunita/AFX-…

#AFXResearch update - added:

Comparative Study of State-based Neural Networks for Virtual Analog Audio Effects Modeling

📜: arxiv.org/abs/2405.04124
🔗: mcomunita.github.io/AFX-Research
⭐️🙏: github.com/mcomunita/AFX-…
Marco Pasini (@marco_ppasini) 's Twitter Profile Photo

✨ Train language models directly on continuous data - without tokenization ✨ We propose an easy way to train GPT-style autoregressive models on continuous data, without error accumulation. We test it on audio 🔊, but this method can easily work with other modalities 🎆 👇🧵

✨ Train language models directly on continuous data - without tokenization ✨

We propose an easy way to train GPT-style autoregressive models on continuous data, without error accumulation.

We test it on audio 🔊, but this method can easily work with other modalities 🎆

👇🧵
Marco Comunità (@marcomunita) 's Twitter Profile Photo

Help us with our research to: 👍 - Develop a perceptual similarity metric for audio effects 🤘 - Advance the state of the art in audio effects modelling Take our listening test (<15min): mcomunita.github.io/mushra-fronten… Use: 💻 + 🎧 🙏

Marco Comunità (@marcomunita) 's Twitter Profile Photo

I admire bands that are brave enough to release one banger of an album and then f*** off...so punk... open.spotify.com/album/20mATsqa…

Marco Comunità (@marcomunita) 's Twitter Profile Photo

#AFXResearch update - added: Hybrid Neural Audio Effects 📜: smcnetwork.org/smc2024/papers… 🔗: mcomunita.github.io/AFX-Research ⭐️🙏: github.com/mcomunita/AFX-…

Eleonora Grassucci (@eleonoragrassuc) 's Twitter Profile Photo

No more pairwise cosine similarity for multimodal learning!🔴 👉GRAM can align *from 2 to n modalities* altogether in a joint fashion by gaining alignment insights from the volume of the parallelotope spanned by the modality vectors. More modalities, better performance! 1/n🧵

Emilian Postolache (@emilianpostola1) 's Twitter Profile Photo

Check out our new work on 📽️V2A🔊! RMS-Mapper with Optical Flow + Stable Audio Open ControlNet = SOTA 🔥 Thanks a lot to the two amazing leads Riccardo Fosco Gramaccioni and Christian Marinoni and to our great collaborators Marco Comunità, Luca Cosmo, Joshua Reiss, and Danilo Comminiello!

Marco Comunità (@marcomunita) 's Twitter Profile Photo

The only Red Hot album I like...for sure thanks to Dave Navarro replacing John Frusciante for this one open.spotify.com/album/74cQBDmk…