探索大语言模型对南亚音乐的理解与生成
Exploring LLMs for South Asian Music Understanding and Generation
AI总结：
本文系统评估大语言模型在基于拉格和塔拉的南亚古典音乐理解与生成任务中的表现，发现前沿模型在理解任务上准确率达85-90%，但生成任务中风格忠实度仅40%。
摘要：
近年来，大语言模型（LLMs）在音乐理解和生成任务中展现出令人瞩目的成果。然而，现有研究仍局限于西方调性传统，未能揭示当前LLMs能否处理结构独特的低资源音乐传统。我们首次系统评估LLMs在南亚古典音乐中的能力——这种传统由拉格（raga）和塔拉（tala）的旋律约束主导，其结构原则与西方和声驱动音乐根本不同。我们的评估基于印度斯坦古典理论和孟加拉古典形式，包括拉宾德拉（Rabindra）和纳兹鲁尔（Nazrul）歌曲——南亚古典音乐中具有代表性的低资源传统。在音乐理解评估中，我们引入了一个包含504个问答的基准测试，涵盖拉格语法、文化知识和符号记谱推理，评估了33个LLMs，其中前沿模型如Gemini 2.5 Pro达到85-90%的准确率，而大多数开源模型仅在23-40%范围内。在音乐生成方面，我们设计了一个五级受控提示框架，发现即使最强的模型也只有40%的时间能产生风格忠实的输出。这些结果表明，音乐生成中的结构有效性和风格忠实度是不同的目标，并突显了文化基础音乐建模的一个开放挑战。
英文摘要：
Recent advancements in Large Language Models (LLMs) have shown promising results in music understanding and generation tasks. However, existing works remain confined to Western tonal traditions, offering little insight into whether current LLMs can handle structurally distinct low-resource musical traditions. We present the first systematic evaluation of LLM competence in South Asian classical music, a tradition governed by raga, tala-based melodic constraints that impose fundamentally different structural principles from Western harmony-driven music. We ground our evaluation in Hindustani classical theory and Bengali classical forms, including Rabindra and Nazrul Sangeet -- representative low-resource traditions within South Asian classical music. For music understanding evaluation, we introduce a 504-question-answer benchmark spanning raga grammar, cultural knowledge, and symbolic notation reasoning, evaluating 33 LLMs where frontier models such as Gemini 2.5 Pro achieve 85-90% accuracy, while most open-source models remain in the 23-40% range. For music generation, we design a five-level controlled prompting framework and find that even the strongest model produces stylistically faithful outputs only 40% of the time. These results reveal that structural validity and stylistic faithfulness in music generation are distinct objectives and highlight an open challenge for culturally grounded music modeling.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
