arXiv cs.CL· Chi-En Dai, Chih-Pin Tan, Yi-Hsuan Yang·· 13 小时前AI 评分21
大语言模型通过文本与多模态上下文学习预测音乐效价与唤醒度
Predicting Music Valence and Arousal via Textual and Multimodal In-Context Learning in Large Language Models
AI 导读
研究用大语言模型的上下文学习（ICL）改进连续音乐情绪预测，不直接输出分数，而是以辅助数据集的示例修正小型任务专用基础模型的初始预测。基于歌词与生成音频描述的文本 ICL 在 80 shots 下将效价 R² 从 0.110 提升至 0.427，融合原始音频的多模态版本仅用 10 shots 即达 0.346。MER60 数据集全程作为留出测试集，表明免训练 ICL 可作为推理阶段策略。
正文 · 原文
Abstract:While arousal can be reliably predicted from musical audio, emotional valence remains difficult to model due to its subjective and context-dependent nature. In this work, we investigate the potential of in-context learning (ICL) with large language models to advance continuous valence and arousal prediction. Rather than predicting emotion scores directly, our framework leverages contextual demonstrations from an auxiliary dataset to refine the initial predictions produced by a small, task-specific base model. We systematically evaluate a text-based ICL variant using lyrics and generated audio captions, alongside a multimodal variant that incorporates raw audio signals. Using the entire MER60 dataset strictly as a held-out test set, our experiments demonstrate substantial performance gains, particularly for valence. Text-based ICL improves the valence $R^2$ from 0.110 to 0.427 with 80 demonstration shots, while the multimodal variant achieves 0.346 with only 10 shots. These findings suggest that training-free ICL offers an effective inference-time strategy for current language models to predict continuous music emotion.
| Comments: | Submitted to ICASSP 2027 |
| Subjects: | Sound (cs.SD); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.31202 [cs.SD] |
| (or arXiv:2610.31202v1 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2610.31202 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Chi-En Dai [view email]
[v1]
Thu, 22 Oct 2026 12:49:14 UTC (468 KB)
来源：arXiv cs.CL · arxiv.org