arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.17822eess.AScs.CL
收藏

多源证据融合用于音频问答

Multi-Source Evidence Fusion for Audio Question Answering

Tallinn University of Technology, Estonia(塔林技术大学,爱沙尼亚)

Aivo Olev, Tanel Alumäe

AI总结:

本文提出多源融合框架,利用两个大音频语言模型与25种可靠性分层的声学工具,生成可验证的推理链,提升音频问答的逻辑性和准确性。

摘要:

大型音频语言模型(LALMs)能够回答关于语音、音乐和环境声音的问题,但其内部推理过程往往不透明且难以验证。我们描述了TalTech在Interspeech 2026音频推理挑战中针对Agent Track的解决方案,系统评估推理过程质量,包括事实准确性、逻辑严密性和推理链的完整性。我们的多源集成管道使用两个生成独立观察的LALMs,同时一个独立的文本推理模型会将这些观察与25种按可靠性分层的声学工具的输出进行交叉验证。通过将每个推理步骤都基于显式的、可靠性标记的证据,系统生成密集且可验证的推理链。我们的系统在挑战中排名第一,在推理质量指标上远超其他系统。

英文摘要:

Large audio language models (LALMs) can answer questions about speech, music, and environmental sounds, yet their internal reasoning is largely opaque and difficult to validate. We describe TalTech's solution to the Agent Track of the Interspeech 2026 Audio Reasoning Challenge, in which systems are evaluated on reasoning process quality, specifically the factual accuracy, logical soundness, and completeness of their reasoning chains. Our multi-source ensemble pipeline uses two LALMs that generate independent observations, while a separate text-only reasoning model cross-checks these against outputs from 25 acoustic tools organized into reliability tiers. By grounding every inference step in explicit, reliability-tagged evidence, the system produces dense, verifiable reasoning chains. Our system ranked first in the challenge, outperforming all competing systems by a wide margin in challenge's reasoning quality metric.

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。

↑