Your Daily AI Research tl;dr — 2022–08–30 🧠

MULAN: Linking Music and Text, Music separation and OpenAI’s approach to aligning AGI.

Welcome to your official daily AI research tl;dr (often with code and news) for AI professionals where I share the most exciting papers I find daily, along with a one-liner summary to help you quickly determine if the article (and code) is worth investigating.

1️⃣ MUSIC SEPARATION ENHANCEMENT WITH GENERATIVE MODELING

“We propose a post-processing model (the Make it Sound Good (MSG) post-processor) to enhance the output of music source separation systems.”

Link to the paper: https://arxiv.org/pdf/2208.12387.pdf

2️⃣ MULAN: A JOINT EMBEDDING OF MUSIC AUDIO AND NATURAL LANGUAGE

MULAN: “a first attempt at a new generation of acoustic models that link music audio directly to unconstrained natural language music descriptions.”

“Human listeners prefer source estimates of bass and drums that have been post-processed by MSG.”

Link to the paper: https://arxiv.org/pdf/2208.12415.pdf