1 code implementation • 29 Jul 2023 • Igor Pereira, Felipe Araújo, Filip Korzeniowski, Richard Vogl
To facilitate the adoption of this dataset, we publish an easy-to-use Python library to download, process and use MoisesDB.
1 code implementation • 7 Oct 2022 • Matthew C. McCallum, Filip Korzeniowski, Sergio Oramas, Fabien Gouyon, Andreas F. Ehmann
We find that restricting the domain of the pre-training dataset to music allows for training with smaller batch sizes while achieving state-of-the-art in unsupervised learning -- and in some cases, supervised learning -- for music understanding.
1 code implementation • 30 Jul 2021 • Filip Korzeniowski, Sergio Oramas, Fabien Gouyon
Artist similarity plays an important role in organizing, understanding, and subsequently, facilitating discovery in large collections of music.
1 code implementation • 22 Oct 2020 • Filip Korzeniowski, Oriol Nieto, Matthew McCallum, Minz Won, Sergio Oramas, Erik Schmidt
The mood of a song is a highly relevant feature for exploration and recommendation in large collections of music.
no code implementations • 16 Aug 2018 • Filip Korzeniowski, Gerhard Widmer
Common temporal models for automatic chord recognition model chord changes on a frame-wise basis.
no code implementations • 16 Aug 2018 • Filip Korzeniowski, Gerhard Widmer
Chord recognition systems typically comprise an acoustic model that predicts chords for each audio frame, and a temporal model that casts these predictions into labelled chord segments.
no code implementations • 16 Aug 2018 • Filip Korzeniowski, Gerhard Widmer
Finally, we investigate the model's performance on short excerpts of audio.
no code implementations • 5 Apr 2018 • Filip Korzeniowski, David R. W. Sears, Gerhard Widmer
We conduct a large-scale study of language models for chord prediction.
no code implementations • 9 Jun 2017 • Filip Korzeniowski, Gerhard Widmer
We present an end-to-end system for musical key estimation, based on a convolutional neural network.
no code implementations • 1 Feb 2017 • Filip Korzeniowski, Gerhard Widmer
Chord recognition systems use temporal models to post-process frame-wise chord preditions from acoustic models.
2 code implementations • 15 Dec 2016 • Rainer Kelz, Matthias Dorfer, Filip Korzeniowski, Sebastian Böck, Andreas Arzt, Gerhard Widmer
In an attempt at exploring the limitations of simple approaches to the task of piano transcription (as usually defined in MIR), we conduct an in-depth analysis of neural network-based framewise transcription.
no code implementations • 15 Dec 2016 • Filip Korzeniowski, Gerhard Widmer
We show that the learned auditory system extracts musically interpretable features, and that the proposed chord recognition system achieves results on par or better than state-of-the-art algorithms.
1 code implementation • 15 Dec 2016 • Filip Korzeniowski, Gerhard Widmer
We explore frame-level audio feature learning for chord recognition using artificial neural networks.