NES Music Database (NES-MDB)

NES-MDB是由加州大学圣地亚哥分校创建的一个大型多乐器音乐数据集，包含数千首为NES音频合成器编写的曲目。数据集为每首歌曲提供了四声部乐谱以及每个声部的动态和音色表现属性。NES-MDB不仅包含音乐创作的语义信息，还详细记录了如何将这些作品转化为细腻的表演。数据集的应用领域主要集中在自动化音乐创作和表演，旨在通过机器学习方法解决音乐创作与表演中的挑战。

NES-MDB is a large-scale multi-instrumental music dataset created by the University of California, San Diego. It contains thousands of tracks composed for the NES audio synthesizer. For each song, the dataset provides four-part musical scores along with dynamic and timbre performance attributes for each musical part. NES-MDB not only contains the semantic information of musical creations, but also elaborately records how to translate these works into nuanced performances. The dataset is primarily applied in the fields of automated music composition and performance, aiming to resolve challenges in music creation and performance through machine learning approaches.