基于偏最小二乘回归的个人化与头相关传递函数

为了通过相对简单的测量得到特定人的与头相关传递函数（HRTF），通过对已测量HRTF的预处理及主元分析和对人体参数的筛选，提出了一种基于偏最小二乘回归（PLSR）的HRTF个人化方法。并进行了客观误差分析和主观声音定位测试。结果表明估计的HRTF与实际测量的HRTF之间不仅均方误差较小，而且在感知方面也没有太大区别；同时由PLSR估计得到的个人化HRTF在水平面上的主观测试定位准确率明显优于非个人化HRTF，也优于由最小二乘回归（LSR）得到的个人化HRTF。

Partial least square regression based head-related transfer function personalization

A method is proposed to obtain individual head related transfer function (HRTF) based on the partial least square regression (PLSR), in which anthropometric parameters are selected by correlation analysis and the original HRTF data are preprocessed before principal component analysis. Then the objective simulation experiment and subjective sound localization experiment are implemented to evaluate the method. Experiments show that the estimated HRTF has not only a small mean square error, but also little perception difference from the measured one；the localization performance of the individual HRTF estimated by the partial least square regression presented in this paper is more accurate than that of the non-individual HRTF obtained from the database and also more accurate than that estimated via the least square regression (LSR).