Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Tongxi, Yu, Yang, Wang, Qing, Qian, Junlang
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2508.01394
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908477641719808
author Wang, Tongxi
Yu, Yang
Wang, Qing
Qian, Junlang
author_facet Wang, Tongxi
Yu, Yang
Wang, Qing
Qian, Junlang
contents Song generation is regarded as the most challenging problem in music AIGC; nonetheless, existing approaches have yet to fully overcome four persistent limitations: controllability, generalizability, perceptual quality, and duration. We argue that these shortcomings stem primarily from the prevailing paradigm of attempting to learn music theory directly from raw audio, a task that remains prohibitively difficult for current models. To address this, we present Bar-level AI Composing Helper (BACH), the first model explicitly designed for song generation through human-editable symbolic scores. BACH introduces a tokenization strategy and a symbolic generative procedure tailored to hierarchical song structure. Consequently, it achieves substantial gains in the efficiency, duration, and perceptual quality of song generation. Experiments demonstrate that BACH, with a small model size, establishes a new SOTA among all publicly reported song generation systems, even surpassing commercial solutions such as Suno. Human evaluations further confirm its superiority across multiple subjective metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2508_01394
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation
Wang, Tongxi
Yu, Yang
Wang, Qing
Qian, Junlang
Sound
Artificial Intelligence
Audio and Speech Processing
Song generation is regarded as the most challenging problem in music AIGC; nonetheless, existing approaches have yet to fully overcome four persistent limitations: controllability, generalizability, perceptual quality, and duration. We argue that these shortcomings stem primarily from the prevailing paradigm of attempting to learn music theory directly from raw audio, a task that remains prohibitively difficult for current models. To address this, we present Bar-level AI Composing Helper (BACH), the first model explicitly designed for song generation through human-editable symbolic scores. BACH introduces a tokenization strategy and a symbolic generative procedure tailored to hierarchical song structure. Consequently, it achieves substantial gains in the efficiency, duration, and perceptual quality of song generation. Experiments demonstrate that BACH, with a small model size, establishes a new SOTA among all publicly reported song generation systems, even surpassing commercial solutions such as Suno. Human evaluations further confirm its superiority across multiple subjective metrics.
title Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation
topic Sound
Artificial Intelligence
Audio and Speech Processing
url https://arxiv.org/abs/2508.01394