cs.SD - 2023-11-06

Combinatorial Hodge Theory in Simplicial Signal Processing – DAFx2023 Lecture Notes

  • paper_url: http://arxiv.org/abs/2311.03469
  • repo_url: None
  • paper_authors: Georg Essl
  • for: 本研讨会讲述了Combinatorial Hodge Theory在 simplicial signal processing 中的应用,尤其是在数字音频效果(DAFx)领域。
  • methods: 本研究使用了Combinatorial Hodge Theory 来分析 simplicial signal processing 中的信号结构,并通过实验证明了其效果。
  • results: 研究发现,Combinatorial Hodge Theory 可以帮助分析和理解 simplicial signal processing 中的信号结构,并提供了一种新的视角来理解这些信号的性质和特征。
    Abstract Lecture notes of a tutorial on Combinatorial Hodge Theory in Simplicial Signal Processing held at international conference for digital audio effects (DAFx-23) in Copenhagen, Denmark.
    摘要 lecture notes of a tutorial on Combinatorial Hodge Theory in Simplicial Signal Processing held at international conference for digital audio effects (DAFx-23) in Copenhagen, Denmark.Translation:DAFx-23国际音频特效会议上的一个教程: combinatorial Hodge theory in simplicial signal processing。

A Foundation Model for Music Informatics

  • paper_url: http://arxiv.org/abs/2311.03318
  • repo_url: https://github.com/minzwon/musicfm
  • paper_authors: Minz Won, Yun-Ning Hung, Duc Le
  • for: 本研究探讨了适用于音乐信息学领域的基础模型,这个领域受到标注数据稀缺和泛化问题的限制。
  • methods: 我们进行了多种基础模型变体的比较研究,检查了关键的决定因素,包括模型架构、tokenization方法、时间分辨率、数据和模型可扩展性。
  • results: 我们的研究发现,我们的模型在多种音乐信息检索下表现出色,在特定的关键指标上超越了现有模型。这些发现对自动学习在音乐信息学中的理解做出了贡献,并为开发更有效和多样的基础模型提供了道路。我们公开发布了一个预训练的版本我们的模型,以便促进重现和未来研究。
    Abstract This paper investigates foundation models tailored for music informatics, a domain currently challenged by the scarcity of labeled data and generalization issues. To this end, we conduct an in-depth comparative study among various foundation model variants, examining key determinants such as model architectures, tokenization methods, temporal resolution, data, and model scalability. This research aims to bridge the existing knowledge gap by elucidating how these individual factors contribute to the success of foundation models in music informatics. Employing a careful evaluation framework, we assess the performance of these models across diverse downstream tasks in music information retrieval, with a particular focus on token-level and sequence-level classification. Our results reveal that our model demonstrates robust performance, surpassing existing models in specific key metrics. These findings contribute to the understanding of self-supervised learning in music informatics and pave the way for developing more effective and versatile foundation models in the field. A pretrained version of our model is publicly available to foster reproducibility and future research.
    摘要 中文翻译:这篇论文研究了针对音乐信息学领域特有的基础模型,该领域面临有限的标注数据和泛化问题。为解决这些问题,我们进行了详细的基础模型变种比较研究,检查了模型结构、字符串化方法、时间分辨率、数据和模型缩放等因素的影响。我们的目标是填补现有知识空白,了解这些因素对音乐信息学中基础模型的成功的贡献。我们采用了严格的评价框架,评估这些模型在音乐信息检索下涉及多个下游任务中的表现,尤其是字符级别和序列级别分类。我们的结果表明,我们的模型在特定的关键指标上达到了稳定的表现,超过了现有模型。这些成果对音乐信息学中自动学习的理解和未来研究做出了贡献。我们还公开提供了我们模型的预训练版本,以便促进可重复性和未来研究。