arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.04029cs.CL
收藏

SpeakerSleuth: 大型音频-语言模型能否在多轮对话中判断说话者一致性?

SpeakerSleuth: Can Large Audio-Language Models Judge Speaker Consistency across Multi-turn Dialogues?

Graduate School of Data Science, Seoul National University(首尔国立大学数据科学研究生院)

Jonggeun Lee, Junseong Pyo, Gyuhyeon Seo, Yohan Jo

AI总结:

本文提出SpeakerSleuth基准测试,评估大型音频-语言模型在多轮对话中判断说话者一致性的能力,发现模型在识别声音不一致性和处理文本上下文时存在显著偏差。

摘要:

大型音频-语言模型(LALMs)作为评估语音生成质量的评判者已崭露头角,但其在多轮对话中评估说话者一致性的能力仍待探索。我们提出了SpeakerSleuth基准测试,通过三个任务评估LALMs是否能可靠地判断多轮对话中的说话者一致性。我们构建了1,818个由人类验证的评估实例,覆盖四个多样化的数据集,包含合成和真实语音,并具有受控的声学难度。评估十二种广泛使用的LALMs后发现,模型在可靠检测声音不一致性方面存在困难。例如,对于同一说话者回合的音频样本,一些模型过度预测不一致性,而另一些模型则过于宽松。模型进一步难以识别出具体的问题回合。当其他对话参与者的回合提供文本上下文时,性能急剧下降,因为模型优先考虑文本连贯性而非声学线索,甚至无法检测到明显的性别转换。另一方面,模型在比较和排序声学变体时表现显著更好,显示出内在的声学辨别能力。这些发现揭示了LALMs存在显著的偏见:它们倾向于优先考虑文本而非声学,暴露了根本的模态不平衡问题,需要解决以构建可靠的音频-语言评判者。我们的代码和数据可在https://github.com/holi-lab/SpeakerSleuth获取。

英文摘要:

Large Audio-Language Models (LALMs) as judges have emerged as a prominent approach for evaluating speech generation quality, yet their ability to assess speaker consistency across multi-turn dialogues remains unexplored. We present \textbf{SpeakerSleuth}, a benchmark evaluating whether LALMs can reliably judge speaker consistency across multi-turn dialogues through three tasks reflecting real-world requirements. We construct 1,818 human-verified evaluation instances across four diverse datasets spanning synthetic and real speech, with controlled acoustic difficulty. Evaluating twelve widely-used LALMs, we find that models struggle to reliably detect acoustic inconsistencies. For instance, given audio samples of the same speaker's turns, some models overpredict inconsistency, whereas others are overly lenient. Models further struggle to identify the exact turns that are problematic. When other interlocutors' turns are provided as textual context, performance degrades dramatically as models prioritize textual coherence over acoustic cues, failing to detect even obvious gender switches for a speaker. On the other hand, models perform substantially better in comparing and ranking acoustic variants, demonstrating inherent acoustic discrimination capabilities. These findings expose a significant bias in LALMs: they tend to prioritize text over acoustics, revealing fundamental modality imbalances that need to be addressed to build reliable audio-language judges. Our code and data are available at https://github.com/holi-lab/SpeakerSleuth.

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。

↑