A personalized head related transfer function generation system and method based on sectional type multiple linear regression

Abstract

The invention provides a personalized head related transfer function generation system and method based on sectional type multiple linear regression. When a function is generated, time-delay removing, redundancy information simplification and data grouping processing are carried out on head-related impulse response data in a sample head-related transfer function information base. Then the grouped data and anthropometric parameter data in the sample database are subjected to multivariate linear regression analyzing to establish a personalized model. In personalization calculation, by inputting the anthropometric parameter information of a personalized object, the head-related pulse response data corresponding to various directions are obtained through personalization model operation; and fine adjustment of anthropometric parameters is carried out to finally generate the personalized head-related transfer function suitable to a determined object. The personalized head related transfer function generation system and method of the invention directly process head-related impulse responses in the head-related transfer function sample base without frequency domain analysis of the data, and has the characteristics of high operation efficiency, high data compression rate and an obvious personalization effect.

Description

technical field [0001] The invention relates to the field of three-dimensional audio effect enhancement, in particular to the establishment of a personalized head-related transfer function model suitable for binaural earphone equipment and a method for using the model to generate a personalized head-related function matched with a user. Background technique [0002] With the gradual development of virtual reality (VR) technology, traditional stereo sound effects can no longer meet the requirements of immersive sound experience, and multi-channel 3D audio technology is very difficult due to expensive playback equipment and complex deployment requirements. Difficult to enter the field of personal use. [0003] Head Related Transfer Function (Head Related Transfer Function, HRTF for short) describes the changes that occur when the sound is transmitted from a specific direction into the human ear through reflection and diffraction of the auricle, torso and other human body parts...

AI Technical Summary

Problems solved by technology

Obtaining HRTF data by experimental measurement requires high requirements on the equipment and environment required for the experiment, and the measurement process is very complicated, which is difficult for individual users to implement

However, for establishing a human body parameter model and obtaining HRTF through calculation, there are disadvantages such as difficult modeling, complex calculation, and low accuracy.

Images