Amadeus: Autoregressive Model with Bidirectional Attribute Modelling for Symbolic Music

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Su, Hongju, Li, Ke, Yang, Lan, Zhang, Honggang, Song, Yi-Zhe
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915468100501504
author Su, Hongju
Li, Ke
Yang, Lan
Zhang, Honggang
Song, Yi-Zhe
author_facet Su, Hongju
Li, Ke
Yang, Lan
Zhang, Honggang
Song, Yi-Zhe
contents Existing state-of-the-art symbolic music generation models predominantly adopt autoregressive or hierarchical autoregressive architectures, modelling symbolic music as a sequence of attribute tokens with unidirectional temporal dependencies, under the assumption of a fixed, strict dependency structure among these attributes. However, we observe that using different attributes as the initial token in these models leads to comparable performance. This suggests that the attributes of a musical note are, in essence, a concurrent and unordered set, rather than a temporally dependent sequence. Based on this insight, we introduce Amadeus, a novel symbolic music generation framework. Amadeus adopts a two-level architecture: an autoregressive model for note sequences and a bidirectional discrete diffusion model for attributes. To enhance performance, we propose Music Latent Space Discriminability Enhancement Strategy(MLSDES), incorporating contrastive learning constraints that amplify discriminability of intermediate music representations. The Conditional Information Enhancement Module (CIEM) simultaneously strengthens note latent vector representation via attention mechanisms, enabling more precise note decoding. We conduct extensive experiments on unconditional and text-conditioned generation tasks. Amadeus significantly outperforms SOTA models across multiple metrics while achieving at least 4$\times$ speed-up. Furthermore, we demonstrate training-free, fine-grained note attribute control feasibility using our model. To explore the upper performance bound of the Amadeus architecture, we compile the largest open-source symbolic music dataset to date, AMD (Amadeus MIDI Dataset), supporting both pre-training and fine-tuning.
format Preprint
id arxiv_https___arxiv_org_abs_2508_20665
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Amadeus: Autoregressive Model with Bidirectional Attribute Modelling for Symbolic Music
Su, Hongju
Li, Ke
Yang, Lan
Zhang, Honggang
Song, Yi-Zhe
Sound
Artificial Intelligence
Multimedia
Existing state-of-the-art symbolic music generation models predominantly adopt autoregressive or hierarchical autoregressive architectures, modelling symbolic music as a sequence of attribute tokens with unidirectional temporal dependencies, under the assumption of a fixed, strict dependency structure among these attributes. However, we observe that using different attributes as the initial token in these models leads to comparable performance. This suggests that the attributes of a musical note are, in essence, a concurrent and unordered set, rather than a temporally dependent sequence. Based on this insight, we introduce Amadeus, a novel symbolic music generation framework. Amadeus adopts a two-level architecture: an autoregressive model for note sequences and a bidirectional discrete diffusion model for attributes. To enhance performance, we propose Music Latent Space Discriminability Enhancement Strategy(MLSDES), incorporating contrastive learning constraints that amplify discriminability of intermediate music representations. The Conditional Information Enhancement Module (CIEM) simultaneously strengthens note latent vector representation via attention mechanisms, enabling more precise note decoding. We conduct extensive experiments on unconditional and text-conditioned generation tasks. Amadeus significantly outperforms SOTA models across multiple metrics while achieving at least 4$\times$ speed-up. Furthermore, we demonstrate training-free, fine-grained note attribute control feasibility using our model. To explore the upper performance bound of the Amadeus architecture, we compile the largest open-source symbolic music dataset to date, AMD (Amadeus MIDI Dataset), supporting both pre-training and fine-tuning.
title Amadeus: Autoregressive Model with Bidirectional Attribute Modelling for Symbolic Music
topic Sound
Artificial Intelligence
Multimedia
url https://arxiv.org/abs/2508.20665