arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.01424cs.SDcs.LGeess.AS
收藏

鲁棒自动钢琴转谱的数据驱动分析

A Data-Driven Analysis of Robust Automatic Piano Transcription

Queen Mary University of London(伦敦玛丽女王大学);Yamaha Corporation(雅马哈公司)

Drew Edwards, Simon Dixon, Emmanouil Benetos, Akira Maezawa, Yuta Kusaka

AI总结:

本研究从训练数据视角探究自动钢琴转谱系统,发现模型易过拟合训练数据的声学特性,通过制作MAESTRO新音频并结合数据增强技术,在零样本适配MAPS数据集时取得88.4 F1-score的最优音符起始点准确率,并通过消融实验分析增强技术的作用。

摘要:

近年来,得益于新数据集与建模技术的发展,自动钢琴转谱算法的性能得到了显著提升。现有研究进展主要聚焦于适配Transformer、Perceiver等新型神经网络架构,以构建更精准的转谱系统。本研究从训练数据的视角出发研究转谱系统,通过测试模型在分布外带标注钢琴数据上的性能,揭示了这类模型会严重过拟合训练数据声学特性的问题。我们为MAESTRO数据集制作了一组新音频,该音频在专业录音棚环境中通过雅马哈Disklavier钢琴自动回放录制。在使用MAESTRO数据集的原始版本与重新演奏版本训练时,结合多种数据增强技术,我们在未见过MAPS数据集任何训练数据的情况下,在该数据集上取得了88.4 F1-score的当前最优音符起始点准确率。随后我们通过一系列消融实验分析这些数据增强技术,以更深入地理解其对最终模型的影响。

英文摘要:

Algorithms for automatic piano transcription have improved dramatically in recent years due to new datasets and modeling techniques. Recent developments have focused primarily on adapting new neural network architectures, such as the Transformer and Perceiver, in order to yield more accurate systems. In this work, we study transcription systems from the perspective of their training data. By measuring their performance on out-of-distribution annotated piano data, we show how these models can severely overfit to acoustic properties of the training data. We create a new set of audio for the MAESTRO dataset, captured automatically in a professional studio recording environment via Yamaha Disklavier playback. Using various data augmentation techniques when training with the original and re-performed versions of the MAESTRO dataset, we achieve state-of-the-art note-onset accuracy of 88.4 F1-score on the MAPS dataset, without seeing any of its training data. We subsequently analyze these data augmentation techniques in a series of ablation studies to better understand their influence on the resulting models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑