Pernet Cyril R, Belin Pascal
Brain Research Imaging Centre, Scottish Imaging Network - A Platform for Scientific Excellence Collaboration, University of Edinburgh Edinburgh, UK.
Front Psychol. 2012 Feb 3;3:23. doi: 10.3389/fpsyg.2012.00023. eCollection 2012.
Voice gender perception can be thought of as a mixture of low-level perceptual feature extraction and higher-level cognitive processes. Although it seems apparent that voice gender perception would rely on low-level pitch analysis, many lines of research suggest that this is not the case. Indeed, voice gender perception has been shown to rely on timbre perception and to be categorical, i.e., to depend on accessing a gender model or representation. Here, we used a unique combination of acoustic stimulus manipulation and mathematical modeling of human categorization performances to determine the relative contribution of pitch and timbre to this process. Contrary to the idea that voice gender perception relies on timber only, we demonstrate that voice gender categorization can be performed using pitch only but more importantly that pitch is used only when timber information is ambiguous (i.e., for more androgynous voices).
语音性别感知可以被看作是低级感知特征提取和高级认知过程的混合。尽管语音性别感知显然依赖于低级音高分析，但许多研究表明情况并非如此。事实上，语音性别感知已被证明依赖于音色感知，并且是分类性的，即依赖于访问性别模型或表征。在这里，我们使用了声学刺激操纵和人类分类表现的数学建模的独特组合，以确定音高和音色在此过程中的相对贡献。与语音性别感知仅依赖于音色的观点相反，我们证明语音性别分类可以仅使用音高来进行，但更重要的是，仅当音色信息模糊时（即对于更具中性的声音）才使用音高。
