arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.04463cs.SDeess.AS
收藏

基于文本大语言模型的自监督语音模型评估

Evaluating Self-Supervised Speech Models via Text-Based LLMS

LY Corporation(LY公司);Carnegie Mellon University(卡内基梅隆大学)

Takashi Maekaku, Keita Goto, Jinchuan Tian, Yusuke Shinohara, Shinji Watanabe

AI总结:

该研究提出用文本大语言模型评估自监督语音模型的新指标,通过输入离散令牌与领域提示得到平均对数似然评分,无需额外训练调优,且评分与语音识别任务相关,LLM还能提供说话人验证可用的推理嵌入。

摘要:

自监督学习(Self-Supervised Learning, SSL)因能以低标注成本学习丰富表征、适配多种下游任务而备受关注。但由于额外训练和评估的成本较高,评估其下游任务性能仍具挑战性。现有的任务无关评估方法也需要额外训练或超参数调优。我们提出一种使用大语言模型(large language models, LLMs)的新型评估指标。将SSL模型生成的离散令牌序列与少量领域提示输入LLM,得到平均对数似然值;这些提示可引导上下文学习,无需额外训练或超参数调优即可让评分更可靠。实验结果表明,基于LLM的评分与自动语音识别任务性能存在相关性。此外,研究发现LLM不仅可作为SSL评估工具,还能提供可用于说话人验证任务的推理时嵌入。

英文摘要:

Self-Supervised Learning (SSL) has gained traction for its ability to learn rich representations with low labeling costs, applicable across diverse downstream tasks. However, assessing the downstream-task performance remains challenging due to the cost of extra training and evaluation. Existing methods for task-agnostic evaluation also require extra training or hyperparameter tuning. We propose a novel evaluation metric using large language models (LLMs). By inputting discrete token sequences and minimal domain cues derived from SSL models into LLMs, we obtain the mean log-likelihood; these cues guide in-context learning, rendering the score more reliable without extra training or hyperparameter tuning. Experimental results show a correlation between LLM-based scores and automatic speech recognition task. Additionally, our findings reveal that LLMs not only functions as an SSL evaluation tools but also provides inference-time embeddings that are useful for speaker verification task.

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。

↑