Self-supervised learning (SSL) has recently emerged as a promising paradigm for training generalisable models on large-scale data in the fields of vision, text, and speech. Although SSL has been proven effective in speech and audio, its application to music audio has yet to be thoroughly explored. This is partially due to the distinctive challenges associated with modelling musical knowledge, particularly tonal and pitched characteristics of music. To address this research gap, we propose an acoustic Music undERstanding model with large-scale self-supervised Training (MERT), which incorporates teacher models to provide pseudo labels in the masked language modelling (MLM) style acoustic pre-training. In our exploration, we identified an effective combination of teacher models, which outperforms conventional speech and audio approaches in terms of performance. This combination includes an acoustic teacher based on Residual Vector Quantisation - Variational AutoEncoder (RVQ-VAE) and a musical teacher based on the Constant-Q Transform (CQT). Furthermore, we explore a wide range of settings to overcome the instability in acoustic language model pre-training, which allows our designed paradigm to scale from 95M to 330M parameters. Experimental results indicate that our model can generalise and perform well on 14 music understanding tasks and attain state-of-the-art (SOTA) overall scores.

翻译：自监督学习（SSL）近期在视觉、文本和语音等领域的大规模数据通用模型训练中展现出广阔前景。尽管SSL在语音和音频领域已被证实有效，但其在音乐音频中的应用尚未得到充分探索，部分原因在于音乐知识建模（尤其是音调与音高特征）面临独特挑战。为填补这一研究空白，我们提出了一种基于大规模自监督训练的声学音乐理解模型（MERT），该模型通过引入教师模型，以掩码语言建模（MLM）风格的声学预训练方式生成伪标签。经过探索，我们确定了一种有效的教师模型组合，其性能优于传统语音与音频方法。该组合包含基于残差向量量化-变分自编码器（RVQ-VAE）的声学教师模型和基于常数Q变换（CQT）的音乐教师模型。此外，我们广泛探索了多种设置以克服声学语言模型预训练中的不稳定性，从而将所设计范式从9500万参数扩展至3.3亿参数。实验结果表明，我们的模型可泛化至14项音乐理解任务并取得优异表现，整体得分达到当前最优（SOTA）水平。