Appropriate Voices for Artefacts: Some Key Insights
Appropriate Voices for Artefacts: Some Key Insights
复制标题
文物的适当声音：一些关键见解
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
Roger K. Moore
中科院分区：
文献类型：
--
作者：
Roger K. Moore
2011年Siri的发布标志着语音技术在能力和可用性方面取得了令人印象深刻的持续进步。随后几年，Google Now等竞争对手出现，亚马逊Echo等消费产品紧随其后。这些设备被视为迈向更先进的“对话”人工制品（尤其是机器人）的第一步。然而，有证据表明，这种语音设备的使用率低得惊人，这可能是由于环境中的噪音、隐私问题或手动替代方案。另一个可能的影响因素是，无处不在的非生命人工制品的不适当的人类声音的部署可能会欺骗用户高估他们的能力，从而造成期望的冲突，最终导致沟通的崩溃。本文强调了基于三项独立研究为给定人工制品提供适当声音的贝内。结果表明，一个非人类的声音的积极影响，并说明如何“适当性”可能会被客观地衡量。最后，给出了一个为MiRo仿生机器人实现适当语音的工作实例。它的结论是，这些见解可能是重要的未来几代的语音人工制品的设计。
The 2011 release of Siri hailed the beginning of a sustained period of impressive advances in the capability and availability of spoken language technology. Subsequent years saw the appearance of competitors such as Google Now , swiftly followed by consumer products such as Amazon Echo . These devices are seen as the first steps towards more advanced ‘conversational’ artefacts (especially robots ). However, evidence suggests that the usage of such voice-enabled devices is surprisingly low, perhaps due to noise in the environment, privacy concerns or manual alternatives.. Another possible contributing factor is that the ubiquitous deployment of inappropriate humanlike voices for non-living artefacts might deceive users into overestimating their capabilities, thereby creating a conflict of expectations that ultimately leads to a breakdown in communications. This paper highlights the benefits of providing an appropriate voice for a given artefact based on three separate studies. Results are presented that demonstrate the positive impact of a non-human voice and illustrate how ‘appropriateness’ might be measured objectively. Finally, a worked-example is presented of implementing an appropriate voice for the MiRo biomimetic robot. It is concluded that these insights could be important for the design of future generations of voice-enabled artefacts.