EvA: An Evidence-First Audio Understanding Paradigm for LALMs 文章

ArXiv CS.AI2026-05-29NEWSen作者: Xinyuan Xie, Shunian Chen, Zhiheng Liu, Yuhao Zhang, Zhiqiang Lv, Liyin Liang, Benyou Wang
查看原文 →
关系图谱

EvA: An Evidence-First Audio Understanding Paradigm for LALMs · 相关技术

相关技术

evidence bottleneckLarge Audio Language Model