Happy to announce that my paper "Adversarial Permutation Invariant Training for Universal Sound Separation" done at Dolby with Jordi Pons, Santi PdP and Joan Serrà has been accepted at IEEE ICASSP #ICASSP2023. 🎉🎉

I'm excited to introduce you Camoscio: an Italian instruction-tuned LLaMA, following Stanford Alpaca. The model should provide output of similar quality to GPT text-davinci-003 and has been finetuned by translating the Alpaca dataset to Italian. github.com/teelinsan/camo… 1/3

Was fortunate enough to be hosted by RAI Radio3 Scienza to talk about GPT4, hallucinations and ethics -- A fun and unique opportunity to engage in thought-provoking discussions with a broader but nonetheless brilliant audience!

Thrilled to share this news made possible by this incredible team! 🚀

Heading to Sicily with our hydrophone to collect data and celebrate World Ocean Passive Acoustic Monitoring Day! IMATI CNR Michele Mancusi Emanuele Rodolà

Check out our latest paper on fast machine translation without re-training! Accepted at ACL-2023 main track 🥳 Amazing collaboration with Andrea Santilli Silvio Severino Valentino Maiorca, Michele Mancusi Riccardo Marin and Emanuele Rodolà

We're excited to announce the first edition of 🔵🔴 UniReps: the Workshop on Unifying Representations in Neural Models! 🧠 To be held at NeurIPS Conference 2023! SUBMISSION DEADLINE: 4 October Check out our Call for Papers, lineup of speakers and schedule at: unireps.org

Proud to present SyncFusion 🥳, a new onset-synchronized foley synthesizer 🔉, done with the amazing Marco Comunità and Riccardo Fosco Gramaccioni! We obtain substantial gains in onset detection 📈 and improved control 🎛️. Big kudos to our advisors Emanuele Rodolà Danilo Comminiello QMUL_IntelSoundEng

Nice to see that our ACL2023 paper on parallel decoding is now extended to speeding up LLMs 😎 Props to LMSYS Org 👏👏

Thrilled to announce that our paper 'MSDM' has been selected for an Oral presentation at #ICLR24, placing in the top 1.2%! Congratulations to the incredible team -Giorgio Mariani, Emilian Postolache , @irenetall25 and our outstanding advisors, Luca Cosmo and Emanuele Rodolà 🎉🎉

Can you reuse neural components without knowing the transformations relating their latent spaces❓🔍 In our #ICLR2024 ✨spotlight✨ paper we show that's indeed possible! From Bricks 🧱..🧱 to Bridges 🌉..🌉: Product of Invariances to Enhance Latent Space Communication [1/4] 🧵

buttered-shrine-021.notion.site/Single-Channel… Your source-separator falls down under one of these categories! 😀 I believe it is a good taxonomy for existing methods! Feel free to comment papers in the relevant category on Notion (I put some of mine as well 😇) and to discuss! [1/3]

#ICLR24 was unforgettable! My first oral presentation at a conference, sharing the stage w/ the great @irenetall25! Thanks again to the other co-authors of MSDM Giorgio Mariani Michele Mancusi and to our advisors Luca Cosmo Emanuele Rodolà! To the next adventure in #musicgen!

🎉 Excited to share that our paper, “COCOLA: Coherence-Oriented Contrastive Learning of Musical Audio Representations,” has been accepted to #ICASSP2025! 🎉 A huge thank-you to my amazing co-authors for their hard work and support. 🙌

🎉Check out our new work on Timbre Transfer, which has been accepted at #ICASSP2025! 🎶 Latent Diffusion Bridges for Unsupervised Musical Audio Timbre Transfer 🔥 arXiv: arxiv.org/abs/2409.06096 Demo: sony.github.io/diffusion-timb… Code: github.com/sony/diffusion… 🙌Thanks to our

🎉 Hyped that 3 of our papers are accepted at #ICASSP2025! Huge congrats to my co-authors, and thanks for their efforts! 👏 - COCOLA: Coherence-Oriented Contrastive Learning of Musical Audio Representations arxiv.org/pdf/2404.16969 - Latent Diffusion Bridges for Unsupervised

🚀Calling ML on Graphs Enthusiasts! I'm co-organizign a Competition on Learning with Noisy Graph Labels at IJCNN 2025. Submissions are now open sites.google.com/view/learning-… Submission Deadline: 10/02/2024 We can’t wait to see your creative approaches and groundbreaking ideas.✨

Moises AI Studio is here. The fastest way to make stems that lock into your track. Forget scrolling samples. Drop in your idea and get drums, bass, guitar, and more that match your sound instantly. 👉Now available on the Web and Desktop App! Try it free at

Huge thanks and congrats to Chiara, Michele Mancusi Donato Crisostomi Luca Cosmo and Emanuele Rodolà! 🎧Demo: giorgioskij.github.io/stage-demo/ 📜Paper: arxiv.org/abs/2504.05690 💻Code: github.com/giorgioskij/st… 🧵(4/4)