Aurian Quélennec

Twenty-Five Years of MIR Research: Achievements, Practices, Evaluations, and Future Challenges Geoffroy Peeters, Zafar Rafii, Magdalena Fuentes, Zhiyao Duan, Emmanouil Benetos, Juhan Nam, Yuki Mitsufuji

A review of 25 years of Music Information Retrieval (MIR) research covers achievements, practices, evaluations, and future challenges; emphasizes signal processing, open research, industry engagement, and diversity.

I released MATPAC++, an audio SSL foundation model with SOTA performances. 🥇

It solves uncertainty in Masked Audio Modelling by selecting the best prediction among several hypotheses. 🧠

📝 Paper: arxiv.org/pdf/2508.12709

👨‍💻 Code & weights soon: github.com/aurianworld/...

Masked Latent Prediction and Classification for Self-Supervised Audio Representation Learning Recently, self-supervised learning methods based on masked latent prediction have proven to encode input data into powerful representations. However, during training, the learned latent space can be f...

You can find the complete paper at the following link:
📝 ieeexplore.ieee.org/document/108...

GitHub - aurianworld/matpac: Official MATPAC implementation and trained model's weights. Official MATPAC implementation and trained model's weights. - aurianworld/matpac

I am excited to share the code from our paper MATPAC, which I will present at ICASSP next week.🇮🇳

This code includes:
⚡Inference code, which you can pip install.
⚙️Full training code for reproducibility.

👨🏾‍💻 github.com/aurianworld/...