用大模型零样本分类音乐流派，效果优于传统方法。
Music Genre Classification using Large Language Models
- 将音频切片后用编码器提取特征，再通过分类头预测流派。
- 音频谱图Transformer（AST）准确率达85.5%，表现最优。
- 适合对音乐信息检索感兴趣的研究者参考。
本文利用预训练大语言模型（LLMs）的零样本能力进行音乐流派分类。该方法将音频信号分割为20毫秒片段，通过卷积特征编码器、Transformer编码器及额外层对音频单元编码并生成特征向量，用于训练分类头。推理时，对各片段的预测结果进行聚合得到最终流派分类。我们全面比较了WavLM、HuBERT、wav2vec 2.0等大模型与1D/2D卷积神经网络（CNN）及音频谱图Transformer（AST）等传统深度学习架构。结果显示，基于Transformer的AST模型整体准确率达到85.5%，优于所有其他评估模型。这表明，即使在零样本场景下，大模型与Transformer架构也具备推动音乐信息检索任务的潜力。
原文摘要 · Abstract (English)
This paper exploits the zero-shot capabilities of pre-trained large language models (LLMs) for music genre classification. The proposed approach splits audio signals into 20 ms chunks and processes them through convolutional feature encoders, a transformer encoder, and additional layers for coding audio units and generating feature vectors. The extracted feature vectors are used to train a classification head. During inference, predictions on individual chunks are aggregated for a final genre classification. We conducted a comprehensive comparison of LLMs, including WavLM, HuBERT, and wav2vec 2.0, with traditional deep learning architectures like 1D and 2D convolutional neural networks (CNNs) and the audio spectrogram transformer (AST). Our findings demonstrate the superior performance of the AST model, achieving an overall accuracy of 85.5%, surpassing all other models evaluated. These results highlight the potential of LLMs and transformer-based architectures for advancing music information retrieval tasks, even in zero-shot scenarios.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。