000165516 001__ 165516
000165516 005__ 20260112132214.0
000165516 0247_ $$2doi$$a10.1371/journal.pone.0335853
000165516 0248_ $$2sideral$$a147251
000165516 037__ $$aART-2025-147251
000165516 041__ $$aeng
000165516 100__ $$aMarcos, Miguel$$uUniversidad de Zaragoza
000165516 245__ $$aRandom rotational embedding Bayesian optimization for human-in-the-loop personalized music generation
000165516 260__ $$c2025
000165516 5060_ $$aAccess copy available to the general public$$fUnrestricted
000165516 5203_ $$aGenerative deep learning models, such as those used for music generation, can produce a wide variety of results based on perturbations of random points in their latent space. User preferences can be incorporated in the generative process by replacing this random sampling with a personalized query. Bayesian optimization, a sample-efficient nonlinear optimization method, is the gold standard for human-in-the-loop optimization problems, such as finding this query. In this paper, we present random rotational embedding Bayesian optimization (ROMBO). This novel method can efficiently sample and optimize high-dimensional spaces with rotational symmetries, like the Gaussian latent spaces found in generative models. ROMBO works by embedding a low-dimensional Gaussian search space into a high-dimensional one through random rotations. Our method outperforms several baselines, including other high-dimensional Bayesian optimization variants. We evaluate our algorithm through a music generation task. Our evaluation includes both simulated experiments and real user feedback. Our results show that ROMBO can perform efficient personalization of a generative deep learning model. The main contributions of our paper are: we introduce a novel embedding strategy for Bayesian optimization in high-dimensional Gaussian sample spaces; achieve a consistently better performance throughout optimization with respect to baselines, with a final loss reduction of 16%-31% in simulation; and complement our simulated evaluations with a study with human volunteers (n = 16). Users working with our music generation pipeline find new favorite pieces 40% more often, 16% faster, and spend 18% less time on pieces they dislike than when randomly querying the model. These results, along with a final survey, demonstrate great performance and satisfaction, even among users with particular tastes. © 2025 Marcos et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
000165516 536__ $$9info:eu-repo/grantAgreement/ES/DGA/T45-23R$$9info:eu-repo/grantAgreement/ES/MICINN-AEI/PID2021-125209OB-I00$$9info:eu-repo/grantAgreement/ES/MICINN/PID2024–158322OB-I00
000165516 540__ $$9info:eu-repo/semantics/openAccess$$aby$$uhttps://creativecommons.org/licenses/by/4.0/deed.es
000165516 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/publishedVersion
000165516 700__ $$aMur-Labadia, Lorenzo$$uUniversidad de Zaragoza
000165516 700__ $$0(orcid)0000-0002-6741-844X$$aMartinez-Cantin, Ruben$$uUniversidad de Zaragoza
000165516 7102_ $$15007$$2520$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Ingen.Sistemas y Automát.
000165516 773__ $$g20, 11 (2025), e0335853 [27 pp.]$$pPLoS One$$tPLoS ONE$$x1932-6203
000165516 8564_ $$s5441462$$uhttps://zaguan.unizar.es/record/165516/files/texto_completo.pdf$$yVersión publicada
000165516 8564_ $$s2096375$$uhttps://zaguan.unizar.es/record/165516/files/texto_completo.jpg?subformat=icon$$xicon$$yVersión publicada
000165516 909CO $$ooai:zaguan.unizar.es:165516$$particulos$$pdriver
000165516 951__ $$a2026-01-12-11:09:03
000165516 980__ $$aARTICLE