Towards Effective Negation Modeling in Joint Audio-Text Models for Music

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Vasilakis, Yannis, Bittner, Rachel, Pauwels, Johan
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917211586691072
author Vasilakis, Yannis
Bittner, Rachel
Pauwels, Johan
author_facet Vasilakis, Yannis
Bittner, Rachel
Pauwels, Johan
contents Joint audio-text models are widely used for music retrieval, yet they struggle with semantic phenomena such as negation. Negation is fundamental for distinguishing the absence (or presence) of musical elements (e.g., "with vocals" vs. "without vocals"), but current systems fail to represent this reliably. In this work, we investigate and mitigate this limitation by training CLAP models from scratch on the Million Song Dataset with LP-MusicCaps-MSD captions. We introduce negation through text augmentation and a dissimilarity-based contrastive loss, designed to explicitly separate original and negated captions in the joint embedding space. To evaluate progress, we propose two protocols that frame negation modeling as retrieval and binary classification tasks. Experiments demonstrate that both methods, individually and combined, improve negation handling while largely preserving retrieval performance.
format Preprint
id arxiv_https___arxiv_org_abs_2601_13931
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Towards Effective Negation Modeling in Joint Audio-Text Models for Music
Vasilakis, Yannis
Bittner, Rachel
Pauwels, Johan
Sound
Information Retrieval
Machine Learning
Joint audio-text models are widely used for music retrieval, yet they struggle with semantic phenomena such as negation. Negation is fundamental for distinguishing the absence (or presence) of musical elements (e.g., "with vocals" vs. "without vocals"), but current systems fail to represent this reliably. In this work, we investigate and mitigate this limitation by training CLAP models from scratch on the Million Song Dataset with LP-MusicCaps-MSD captions. We introduce negation through text augmentation and a dissimilarity-based contrastive loss, designed to explicitly separate original and negated captions in the joint embedding space. To evaluate progress, we propose two protocols that frame negation modeling as retrieval and binary classification tasks. Experiments demonstrate that both methods, individually and combined, improve negation handling while largely preserving retrieval performance.
title Towards Effective Negation Modeling in Joint Audio-Text Models for Music
topic Sound
Information Retrieval
Machine Learning
url https://arxiv.org/abs/2601.13931