symusic: A swift and unified toolkit for symbolic music processing

Yikai Liao (Beijing University of Posts and Telecommunications), Zhongqi Luo (Qiyin Technology Co., Ltd.)*, Yue Wang (China Conservatory of Music), Yujie Wu (殷渝杰),

This paper will be presented virtually at the 11:15 PM - 11:45 PM PST session

Abstract:

This paper introduces symusic, a swift and unified toolkit for symbolic music processing. It provides high-level APIs for symbolic music manipulation, and an optimized IO engine which supports multiple file formats. Symusic addresses key challenges in symbolic music processing by offering exceptional speed, multi-format support, flexible time unit conversion, and comprehensive batch operations. With up to 1000 times faster parsing than existing libraries, symusic is poised to accelerate large-scale music analysis and generation. Code and pre-built package of symusic are available at Github and PyPI.