摘要
利用宽带波束形成器输出合成了HRTF数据。确立了波束形成器结构,建立了HRTF合成的最优问题,采用二阶锥规划最优方法得到了波束形成器的最优权向量,进而利用波束形成器的输出得到HRTF合成数据。同时结合MIT实验室的HRTF测量数据的基本特性,在求解最优问题的过程中利用了HRTF的最小相位表示,进一步降低了滤波器阶次,简化了波束形成器结构。计算机仿真和主观听觉定位实验表明,在一定频段内该方法得到的HRTF合成近似值和测量数据具有较好的一致性,验证了该方法的有效性。
The head-related transfer function(HRTF) is synthesized using the output of a broadband beamformer.The beamformer structure is determined and the optimization problem of the HRTF synthesis is established.The optimal weighting vector of the beamformer is derived using second-order cone programming(SOCP).Then,the HRTF synthesis data are obtained from the output of the beamfomer.Considering the property of the KEMAR HRTF measurement database from MIT media lab,the mini-phase expression of the HRTF is used in the process of solving the optimization problem.The order of the filter is reduced and the beamformer structure is simplified.Computer simulations and subjective listening experiments demonstrate that the HRTF approximation data are similar to those of the measurements from MIT media lab in certain frequency band,which verifies the validity of the proposed synthesis method.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2011年第11期2455-2460,共6页
Chinese Journal of Scientific Instrument
关键词
波束形成
头相关传递函数
二阶锥规划
虚拟听觉
beamforming
head-related transfer function(HRTF)
second-order cone programming(SOCP)
virtual audition