声学压力下的语音代理:从信号退化到交互与行动
Voice Agents under Acoustic Stress: From Signal Degradation to Interaction and Action
浏览论文内容
中文总结 AI 辅助
本文综述了语音代理在噪声等声学压力下的任务完成能力,提出TRACE工作流用于评估其鲁棒性,并指导改进以防止错误行动。
中文摘要 AI 辅助
语音代理必须在噪声、混响和竞争性语音存在的情况下完成用户的任务。因此,评估代理的鲁棒性需要追踪声学条件如何影响对话以及代表用户采取的行动。本综述考察了现有基准在声学压力下揭示代理完成任务能力的情况,以及哪些方面需要进一步基于任务的评估。随后,我们介绍了TRACE,一个用于设计、运行和解释面向任务的人机交互中声学鲁棒性评估的实用工作流:同一代理使用原始录音和声学压力副本尝试指定任务,并对由此产生的对话进行任务完成、错误行动、恢复和用户努力程度的评分。最后,我们解释了这些评估结果如何指导对代理的修改,以防止错误行动并改善恢复。
英文摘要
Voice agents must complete users' tasks despite noise, reverberation, and competing speech. Evaluating agents' robustness therefore requires following how acoustic conditions affect the conversation and the actions taken on the user's behalf. This overview examines what existing benchmarks reveal about agents' ability to complete tasks under acoustic stress and where further task-based evaluation is required. We then introduce TRACE, a practical workflow for designing, running, and interpreting evaluations of acoustic robustness in task-oriented human-agent interactions: the same agent attempts a specified task with an original recording and an acoustically stressed copy, and the resulting conversations are scored for task completion, wrong actions, recovery, and user effort. Finally, we explain how results from these evaluations can guide changes to an agent to prevent wrong actions and improve recovery.
发表机构
- Technion -- Israel Institute of Technology(以色列理工学院)
- Massachusetts Institute of Technology(麻省理工学院)
- Bar-Ilan University(巴伊兰大学)
- Carnegie Mellon University(卡内基梅隆大学)
机构由 AI 辅助整理,请以论文原文为准。