Symbolic piano music understanding from large-scale pre-training
Symbolic piano music understanding from large-scale pre-training
复制标题
大规模预训练的象征性钢琴音乐理解
DOI：
10.11517/pjsai.jsai2022.0_2s5is2c02
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Hidemoto Nakada
中科院分区：
文献类型：
--
作者：
Yingfeng Fu;Yusuke Tanimura;Hidemoto Nakada
由大量数据驱动的预训练在自然语言理解中显示出巨大的力量。现有的作品使用预训练的符号音乐是不够普遍的，以解决所有的任务，在音乐信息检索。为了弥补这一不足，并与已有的作品进行比较，我们采用了一种类似BERT的掩蔽语言预训练方法，在MAESTRO数据集的复调钢琴曲文件上训练了一个堆叠的Music Transformer [Huang 18]。然后，我们在几个符号音乐理解任务上对预训练模型进行了微调。在我们目前正在进行的工作中，我们补充了几个音符级别的任务，包括下一个标记预测，旋律提取，速度预测和和弦识别。并与前人的研究成果进行了比较。
Pre-training driven by a vast amount of data has shown great power in natural language understanding. The existing works using pretraining for symbolic music are not general enough to tackle all the tasks in musical information retrieval. To make up for the insufficiency and compare it with the existing works, we employed a BERT-like masked language pre-training approach to train a stacked Music Transformer [Huang 18] on polyphonic piano MIDI files from the MAESTRO dataset. Then we finetuned our pre-trained model on several symbolic music understanding tasks. In our current work in progress, we complemented several note-level tasks, including next token prediction, melody extraction, velocity prediction, and chord recognition. And we compared our model with the previous works.