Full-Band Music Genres Interpolations with Wasserstein Autoencoders

Tijn Borghuis, Alessandro Tibo, Simone Conforti, Lorenzo Brusci, Paolo Frasconi

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

2 Downloads (Pure)

Abstract

We compare different types of autoencoders for
generating interpolations between four-instruments
musical patterns in the acid jazz, funk, and soul
genres. Preliminary empirical results suggest the
superiority of Wasserstein autoencoders. The process of generation produces musically sound and
creative transitions between different genres and
can be of interest to practitioners in the field.
Original languageEnglish
Title of host publicationItal-IA
Publication date2019
DOIs
Publication statusPublished - 2019
EventWorkshop AI for Media and Entertainment - Siena, Italy
Duration: 18 Mar 201918 Mar 2019

Conference

ConferenceWorkshop AI for Media and Entertainment
CountryItaly
CitySiena
Period18/03/201918/03/2019

Fingerprint

music
interpolation
acids
acoustics

Cite this

Borghuis, Tijn ; Tibo, Alessandro ; Conforti, Simone ; Brusci, Lorenzo ; Frasconi, Paolo. / Full-Band Music Genres Interpolations with Wasserstein Autoencoders. Ital-IA. 2019.
@inproceedings{b0dd4f98711145b88f3cc9372a29c767,
title = "Full-Band Music Genres Interpolations with Wasserstein Autoencoders",
abstract = "We compare different types of autoencoders forgenerating interpolations between four-instrumentsmusical patterns in the acid jazz, funk, and soulgenres. Preliminary empirical results suggest thesuperiority of Wasserstein autoencoders. The process of generation produces musically sound andcreative transitions between different genres andcan be of interest to practitioners in the field.",
author = "Tijn Borghuis and Alessandro Tibo and Simone Conforti and Lorenzo Brusci and Paolo Frasconi",
year = "2019",
doi = "http://dlw2019.cpr.ku.dk/program/",
language = "English",
booktitle = "Ital-IA",

}

Borghuis, T, Tibo, A, Conforti, S, Brusci, L & Frasconi, P 2019, Full-Band Music Genres Interpolations with Wasserstein Autoencoders. in Ital-IA. Workshop AI for Media and Entertainment, Siena, Italy, 18/03/2019. https://doi.org/http://dlw2019.cpr.ku.dk/program/

Full-Band Music Genres Interpolations with Wasserstein Autoencoders. / Borghuis, Tijn; Tibo, Alessandro; Conforti, Simone; Brusci, Lorenzo; Frasconi, Paolo.

Ital-IA. 2019.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Full-Band Music Genres Interpolations with Wasserstein Autoencoders

AU - Borghuis, Tijn

AU - Tibo, Alessandro

AU - Conforti, Simone

AU - Brusci, Lorenzo

AU - Frasconi, Paolo

PY - 2019

Y1 - 2019

N2 - We compare different types of autoencoders forgenerating interpolations between four-instrumentsmusical patterns in the acid jazz, funk, and soulgenres. Preliminary empirical results suggest thesuperiority of Wasserstein autoencoders. The process of generation produces musically sound andcreative transitions between different genres andcan be of interest to practitioners in the field.

AB - We compare different types of autoencoders forgenerating interpolations between four-instrumentsmusical patterns in the acid jazz, funk, and soulgenres. Preliminary empirical results suggest thesuperiority of Wasserstein autoencoders. The process of generation produces musically sound andcreative transitions between different genres andcan be of interest to practitioners in the field.

U2 - http://dlw2019.cpr.ku.dk/program/

DO - http://dlw2019.cpr.ku.dk/program/

M3 - Article in proceeding

BT - Ital-IA

ER -