roser

Huge thanks to my amazing co-authors: Laura Ibáñez-Martínez, @xavier-serra.bsky.social, Emilia Gómez and
@martinrocamora.bsky.social, and to our colleagues at the
@mtg-upf.bsky.social and the broader MIR community who helped shape this work 🤍

GitHub - roserbatlleroca/MusGO_framework: This repository is complementary to the article "MusGO: A Community-Driven Framework for Asssessing Openness in Music-Generative AI" in Proceedings of the 26t... This repository is complementary to the article "MusGO: A Community-Driven Framework for Asssessing Openness in Music-Generative AI" in Proceedings of the 26th International Society for M...

With it, we assessed openness across 16 music-generative models. But, MusGO isn't just a static paper: it's a live project 🌿

We invite you to explore the full framework, contribute by adding new models, or contesting model evaluations in our repo:
💻 github.com/roserbatller...

We introduce MusGO: a composite and graded framework to evaluate how models share (or don’t share) key components like code, data, weights, documentation, and licensing.

It’s built on 13 categories (8 essential, 5 desirable) and refined with feedback from 100+ MIR community members.

how can we assess openness in music-generative AI?

That’s the question we explore in our new paper, just accepted at #ISMIR2025! 📣

📄 Paper: arxiv.org/pdf/2507.03599

📌 Openness leaderboard: roserbatlleroca.github.io/MusGO_framew...

🧵👇

we're still looking for participants in our latest listening experiment 👂

if you have 20 minutes to spare, it'd be great to get your input on how people perceive sound and recognise similarities between different audio clips 🎶

👇

🎧 participate in our short experiment on sound perception

help us understand how you identify similarities and relationships between sounds 🧠🎶
available both in English and Catalan:
🔗 en - mtg.upf.edu/similarity-e...
🔗 cat - mtg.upf.edu/similarity-e...

thank you for your time and support!

Are you working on music + ethics, perception, UX/UI, human-AI collaboration or exploring future human-centric approaches? Don’t miss out HCMIR workshop call for papers!

More info here👇🏽

📢 New preprint version of "Transparency in Music-Generative AI: A Systematic Literature Review" is out!

We now analyse 107 publications in a two-stage review, uncovering gaps in transparency for genAI in music. Jump to section 5.2 for TL;DR ✨

doi.org/10.21203/rs....

new paper! 🗣️Sketch2Sound💥

Sketch2Sound can create sounds from sonic imitations (i.e., a vocal imitation or a reference sound) via interpretable, time-varying control signals.

paper: arxiv.org/abs/2412.08550
web: hugofloresgarcia.art/sketch2sound

📢 Call to all #MIRCommunity!

Have you ever wondered what an open model means? Help us shape the definition of open models in generative AI for music by taking our survey — just 10 minutes!

👉 forms.gle/Z48t6HPBXwWC3r…

thank you 💫