发表机构
Seoul National University（首尔大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对LALMs缺乏听觉错觉基准的问题，构建了首个含10种错觉的AIB基准，对比模型与人类响应发现LALMs存在系统性差异，为探究听觉认知提供了新测试平台。
AI 中文摘要
感知错觉长期以来都是探究人类认知的重要工具，能够揭示感知的偏差与局限。在听觉领域，这类错觉为测试大型音频语言模型（Large Audio Language Models, LALMs）是否复现人类感知倾向提供了独特视角。尽管听觉错觉十分重要，但大多数基准聚焦于视觉错觉或通用音频任务，使得听觉错觉的相关研究未得到充分探索。为此，我们提出AIB——首个面向LALMs的听觉错觉基准，涵盖音乐、声音、语音领域的10种代表性错觉，每种错觉都标注了是否存在基于知识的先验信息。我们的方法将模型评估与受控的人类听觉研究相结合，以实现对响应的直接对比。结果显示存在系统性差异：多数LALMs在低层级声学错觉上保持对信号的忠实性，而部分模型在涉及语言或音乐先验时表现出更接近人类的响应，不过没有任何模型能匹配人类的感知模式。这些发现凸显了LALMs作为认知模型的当前局限。通过将听觉错觉确立为严格的测试平台，本研究为探究神经黑箱模型、推进对听觉认知的理解提供了新视角。AIB已公开，可访问此httpsURL获取。
英文摘要
Perceptual illusions have long served as crucial probes into human cognition, revealing biases and limitations of perception. In the auditory domain, such illusions provide a unique lens for testing whether Large Audio Language Models (LALMs) replicate human perceptual tendencies. Despite their importance, most benchmarks focus on visual illusions or general audio tasks, leaving auditory illusions underexplored. To this end, we present AIB, the first auditory illusion benchmark for LALMs, covering ten representative illusions across music, sound, and speech, each annotated for the presence of knowledge-based priors. Our methodology pairs model evaluation with controlled human listening studies, enabling direct comparison of responses. Results show systematic differences: while most LALMs remain signal-faithful on low-level acoustic illusions, several exhibit more human-like responses when linguistic or musical priors are involved, although no model matches the human perceptual profile. These findings highlight the current limitations of LALMs as cognitive models. By establishing auditory illusions as a rigorous testbed, our work offers a new perspective for probing neural black-box models and advancing understanding of auditory cognition. AIB is publicly available at https://github.com/gillosae/aib.
CommentsAccepted to the 2026 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2026)
DOI：10.1109/icassp55912.2026.11462423
