TOWARDS STANDARDIZING PERCEPTUAL VOICE QUALITY MEASURES

实现感知语音质量测量的标准化

• 批准号: 6823281
• 负责人: JODY E KREIMAN
• 金额: $ 46.67万
• 依托单位: UNIVERSITY OF CALIFORNIA LOS ANGELES
• 依托单位国家: 美国
• 财政年份: 1992
• 资助国家: 美国
• 起止时间: 1992-12-01 至 2005-11-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6823281
• 关键词: clinical research; computer system design /evaluation; human subject; mathematical model; model design /development; noise; sound perception; speech disorders; speech recognition; speech synthesizers; voice

Judgments of voice quality contribute greatly to patients' opinions of their voice, and also to a clinician's decision to initiate and continue treatment. Quality judgments are further used as a standard against which instrumental measures of voice are validated. Nevertheless, these "subjective" measures of voice quality are not highly regarded as either clinical or research tools, because of problems with reliability and untested validity. We hypothesize that problems in voice quality measurement do not originate within the listener, but rather derive from the methods used to measure what listeners hear. The proposed research departs from these traditional rating scale methods by applying a speech synthesizer for pathological voice quality to study fundamental issues concerning reliability and validity of measures of voice quality. In this approach, listeners an asked to adjust synthesizer parameters to achieve a perceptual match to the original voice -- to "rate" a voice by matching it. This method explicitly links the acoustic signal to perceived voice quality. The investigators hypothesize that this linkage across the speech chain increases the validity, reliability, and utility of both acoustic and perceptual representations, by providing listeners with an objective tool (a synthesizer) for quantifying what they hear. The proposed research focuses on expanding and refining our synthesizer for pathological voice quality, compares the validity and reliability of synthesis techniques to traditional rating scale techniques for evaluating voice quality, applies the synthesizer to examine the perceptual importance of a periodicity, noise and the shape of the voicing source pulse, and quantifies the different sources of error in voice quality measurement. The investigators hypothesize that the protocols described here will greatly increase both the reliability and validity of voice quality measurement, bringing us much closer to our goal to develop a standardized voice evaluation protocol.

对语音质量的判断对患者的病情有很大帮助 他们的声音的意见，以及临床医生决定启动和 继续治疗。质量判断进一步被用作针对的标准 哪些声音的仪器测量是经过验证的。尽管如此，这些 语音质量的“主观”衡量标准也没有受到高度重视 临床或研究工具，由于可靠性问题且未经测试 有效性。我们假设语音质量测量中的问题不会 源自侦听器，而是源自用于侦听的方法 衡量听众听到的内容。拟议的研究偏离了这些 传统的评分量表方法通过应用语音合成器来 病理语音质量研究有关可靠性的基本问题 以及语音质量测量的有效性。在这种方法中，听众会问 调整合成器参数以实现与原始的感知匹配 声音——通过匹配声音来“评价”声音。该方法显式链接 声音信号感知语音质量。研究人员假设 这种跨越语音链的联系提高了有效性、可靠性和 通过向听众提供声音和感知表征的效用 使用客观工具（合成器）来量化他们所听到的内容。这 拟议的研究重点是扩展和完善我们的合成器 病态语音质量，比较合成的有效性和可靠性 用于评估语音质量的传统评级量表技术的技术， 应用合成器来检查周期性的感知重要性， 噪声和发声源脉冲的形状，并量化不同的 语音质量测量中的误差源。研究人员假设 这里描述的协议将大大提高可靠性 和语音质量测量的有效性，使我们更接近我们的目标 开发标准化的语音评估协议。