Audio-Based Musical Version Identification: Elements and challenges

Abstract

This tutorial article provides a review of key ideas and approaches proposed in 20 years of VI research, and to connect them to current practice. For more than a decade, VI systems suffered from the accuracy-scalability trade-off, with attempts to increase accuracy resulting in cumbersome, non-scalable systems. Recent years however have witnessed an increase in VI approaches based on deep learning systems that take a step toward bridging the accuracy-scalability gap, and we start seeing the possibility to deploy such systems in real-world applications. Although this trend positively influences the number of researchers and institutions working on VI, it may also result in obscuring the literature before the deep learning era. To appreciate years of novel ideas in VI and to facilitate building better systems, we believe that now may be the right time to review some of the successful ideas and applications proposed in VI literature and to connect them to current systems.

Related

September 2022 | RecSys

Identifying New Podcasts with High General Appeal Using a Pure Exploration Infinitely-Armed Bandit Strategy

Maryam Aziz, Jesse Anderton, Kevin Jamieson, Alice Wang, Hugues Bouchard, Javed Aslam

September 2022 | Interspeech

Unsupervised Speaker Diarization that is Agnostic to Language Overlap Aware and Free of Tuning

M Iftekhar Tanveer, Diego Casabuena, Jussi Karlgren, Rosie Jones

September 2022 | Interspeech

Exploring audio-based stylistic variation in podcasts

Katariina Martikainen, Jussi Karlgren, Khiet Truong