Audio-based musical version identification: Elements and challenges

F Yesiler, G Doras, RM Bittner… - IEEE Signal …, 2021 - ieeexplore.ieee.org
Creating novel interpretations of existing musical compositions is and has always been an
essential part of musical practice. Before the advent of recorded music, listening to a piece of …

Bytecover: Cover song identification via multi-loss training

X Du, Z Yu, B Zhu, X Chen, Z Ma - ICASSP 2021-2021 IEEE …, 2021 - ieeexplore.ieee.org
We present in this paper ByteCover, which is a new feature learning method for cover song
identification (CSI). Byte-Cover is built based on the classical ResNet model, and two major …

Artist similarity with graph neural networks

F Korzeniowski, S Oramas, F Gouyon - arXiv preprint arXiv:2107.14541, 2021 - arxiv.org
Artist similarity plays an important role in organizing, understanding, and subsequently,
facilitating discovery in large collections of music. In this paper, we present a hybrid …

Training deep pitch-class representations with a multi-label CTC loss

C Weiss, G Peeters - … Society for Music Information Retrieval Conference …, 2021 - hal.science
Despite the success of end-to-end approaches, chroma (or pitch-class) features remain a
useful mid-level representation of music audio recordings due to their direct interpretability …

Deep learning for audio and music

G Peeters, G Richard - Multi-faceted Deep Learning: Models and Data, 2021 - Springer
This chapter provides an overview of how deep learning techniques can be used for audio
signals. We first review the main DNN architectures, meta-architectures and training …

[PDF][PDF] Artist Similarity for Everyone: A Graph Neural Network Approach.

F Korzeniowski, S Oramas… - Trans. Int. Soc. Music. Inf …, 2022 - scholar.archive.org
Artist similarity plays an important role in organizing, understanding, and subsequently,
facilitating discovery in large collections of music. In this paper, we present a hybrid …

Assessing algorithmic biases for musical version identification

F Yesiler, M Miron, J Serrà, E Gómez - … on Web Search and Data Mining, 2022 - dl.acm.org
Version identification (VI) systems now offer accurate and scalable solutions for detecting
different renditions of a musical composition, allowing the use of these systems in industrial …

The deep learning revolution in mir: The pros and cons, the needs and the challenges

G Peeters - International Symposium on Computer Music …, 2019 - Springer
This paper deals with the deep learning revolution in Music Information Research (MIR), ie
the switch from knowledge-driven hand-crafted systems to data-driven deep-learning …

And what if two musical versions don't share melody, harmony, rhythm, or lyrics?

M Abrassart, G Doras - ISMIR 2022, 2022 - hal.science
Version identification (VI) has seen substantial progress over the past few years. On the one
hand, the introduction of the metric learning paradigm has favored the emergence of …

Pairwise similarity normalization based on a hubness score for improving cover song retrieval accuracy

JS Seo - IEICE TRANSACTIONS on Information and Systems, 2022 - search.ieice.org
A hubness-score based normalization of the pairwise similarity is proposed for the sequence-
alignment based cover song retrieval. The hubness, which is the tendency of some data …