跳到主要导航 跳到搜索 跳到主要内容

Symbolic Music Generation with Adaptive Representation Alignment in Texture

  • Qing Chen
  • , Hengyu Zhang
  • , Kaiyuan Liu
  • , Tianyi Lu
  • , Qidang Zhou
  • , Daoguo Dong*
  • , Liang He
  • *此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Symbolic music is gaining more and more attention for its high semantic precision and editorial flexibility. However, using traditional condition processing methods to generate symbolic music is problematic since these approaches often result in semantic loss by overlooking the intrinsic connections between musical elements. We introduce the SMART, Symbolic Music generation with Adaptive Representation alignment in Texture which incorporates a condition disentanglement method and an adaptive representation alignment framework, capable of fully utilizing multi-scale information and generating music with consistent texture. In SMART, the condition disentanglement method is used to decouple complex condition into more refined musical elements and the representation alignment framework is for semantic constraint during the generation process. Our method demonstrates remarkable advancements in texture consistency and chord controllability compared to standard strong baselines in symbolic music generation.

源语言英语
主期刊名McGE 2025 - Proceedings of the 3rd International Workshop on Multimedia Content Generation and Evaluation
主期刊副标题New Methods and Practice, Co-Located with MM 2025
出版商Association for Computing Machinery, Inc
12-18
页数7
ISBN(电子版)9798400720604
DOI
出版状态已出版 - 26 10月 2025
活动3rd International Workshop on Multimedia Content Generation and Evaluation: New Methods and Practice, McGE 2025 - Dublin, 爱尔兰
期限: 31 10月 202531 10月 2025

出版系列

姓名McGE 2025 - Proceedings of the 3rd International Workshop on Multimedia Content Generation and Evaluation: New Methods and Practice, Co-Located with MM 2025

会议

会议3rd International Workshop on Multimedia Content Generation and Evaluation: New Methods and Practice, McGE 2025
国家/地区爱尔兰
Dublin
时期31/10/2531/10/25

指纹

探究 'Symbolic Music Generation with Adaptive Representation Alignment in Texture' 的科研主题。它们共同构成独一无二的指纹。

引用此