Sweaterdog/music-theory-images-20k
收藏官方服务：
资源简介：
该数据集旨在让模型回答关于音乐理论的问题，并训练模型识别图像旋律中的每个音符。乐谱是随机生成的，且设计得尽可能难以阅读。
This dataset is designed to make models answer Q&A about Music Theory. It also trains models to name every note in the melody in the image. The sheet music is randomly generated, it is designed to be as hard to read as possible.
提供机构：
Sweaterdog
