个性化头相关传输函数的近似获取——现状和问题  

Approximation of individualized head-related transfer function——current progresses and problems

作　　者：钟小丽[1] 谢菠荪[1] 

机构地区：[1]华南理工大学理学院物理系声学研究所,广州510641

出　　处：《应用声学》2012年第6期410-415,共6页Journal of Applied Acoustics

基　　金：国家自然科学基金(11004064)资助项目;亚热带建筑科学国家重点实验室自主研究课题

摘　　要：虚拟听觉重放采用头相关传输函数(HRTF)合成双耳声信号,并用耳机重放,以产生所需的空间听觉事件。理想的虚拟听觉重放需要个性化HRTF。个性化HRTF可通过实验测量或数值计算相对地准确获得。然而,测量每个潜在使用者的高空间分辨率HRTF是困难的,而数值计算HRTF的频段往往受限于计算机性能。近年发展了多种HRTF的近似获取方法,并成为热门研究课题,但效果有待验证和提高。本文评述了个性化HRTF近似的研究进展,指出了存在的问题和今后的方向。Virtual auditory display recreates desired spatial auditory events by synthesizing binaural signals by head-related transfer function （HRTF） and subsequently rendering them through headphone. Previous studies have proved that faithful performance of a virtual auditory display needs incorporating individual HRTFs obtained through empirical measurement or numerical computation with a certain accuracy. However, measuring individual HRTFs for each of potential users with high spatial resolution is tiresome and difficult, and available frequency range of numerical computation is severely limited by computer capability. Thus, a variety of approaches to approximating individual HRTFs have been developed, and become hot issues in the research of virtual auditory display. Although a moderate success has been achieved, further validation and improvement of those approximation approaches are required. This paper reviews the latest development in relevant field with emphasis on existing problems and future direction.

关 键 词：虚拟听觉重放 头相关传输函数 个性化 近似 

分 类 号：O428[理学—声学]