Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

HRTF personalization method based on sparse principal component analysis

A sparse principal component and principal component analysis technology, applied in the field of HRTF personalization, can solve problems such as weak local correlation, no consideration of physiological parameter correlation, and no persuasiveness

Active Publication Date: 2019-11-22
NORTHWESTERN POLYTECHNICAL UNIV
View PDF10 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Physiological parameters obtained through experience are not convincing in theory. In contrast, physiological parameters obtained through correlation analysis have a certain theoretical basis. However, this correlation analysis only considers the correlation between physiological parameters. , without considering the correlation between physiological parameters and HRTF
At the same time, the physiological parameters targeted by this method are two-dimensional physiological parameters. The physiological parameters are widely distributed and the local correlation is not strong. For the three-dimensional physiological parameters, due to the relatively dense aggregation, when the correlation is used for dimensionality reduction, physiological parameters will appear. are all correlated, so the method is no longer valid for 3D physiological parameters

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • HRTF personalization method based on sparse principal component analysis
  • HRTF personalization method based on sparse principal component analysis
  • HRTF personalization method based on sparse principal component analysis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] refer to figure 1 . The HRTF personalization process for HRTF with azimuth angle θ=0° and elevation angle φ=0° is introduced in detail.

[0039] Step 1: Select HRTF of all subjects in the orientation (0°, 0°), conduct principal component analysis on the HRTF of this orientation, and obtain the principal component of this orientation:

[0040] (1) Select all subjects in the HRTF at the orientation (0°,0°) to form the vector H ij , where i is the subject sequence, j is the frequency number, and the number of subjects is m, where the azimuth angle θ=0°, the elevation angle φ=0°;

[0041] (2) to H ij Standardize as follows:

[0042]

[0043] in, for H ij Normalized head-related transfer function;

[0044] (3) Perform principal component analysis on the standardized head-related transfer function:

[0045]

[0046] Among them, P m×n is the m×n score matrix of principal component analysis; W is the n×n load matrix of principal component analysis, and T repre...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an HRTF personalization based on sparse principal component analysis. The method comprises the following steps: firstly, performing dimension reduction on HRTF by applying a principal component analysis method; carrying out dimensionality reduction on the physiological parameters by using a sparse principal component analysis method; and finally, taking the physiological parameters subjected to dimension reduction as input, taking the HRTF subjected to dimension reduction as output, and performing nonlinear fitting by applying a generalized regression neural network. According to the method, dimension reduction of three-dimensional physiological parameters can be realized by applying sparse principal component analysis, original dozens of-dimensional physiological parameters are reduced to several dimensions, and the structure of the data is not influenced. Local convergence of data is avoided when a correlation method is applied to dimensionality reduction processing, and the dimensionality reduction effect is improved. And regression is performed on the HRTF principal component coefficient by using the dimension-reduced data, so that the regression time isshortened, and the over-fitting risk is avoided. And HRTF personalization can be better realized by applying the three-dimensional physiological parameters after dimension reduction.

Description

technical field [0001] This patent relates to a HRTF personalization method, especially a HRTF personalization method based on sparse principal component analysis. Background technique [0002] Head Related Transfer Function (Head Related Transfer Function, HRTF) is used to describe the frequency domain acoustic transfer function that the sound emitted by the free field sound source reaches the binaural ears after being scattered and reflected by the head, auricle, trunk and other physiological structures. Each sound source spatial position corresponds to a pair of HRTFs, which are generally functions of the distance from the sound source to the center of the head, the azimuth and elevation angles of the sound source, and the frequency. Because the physiological structure and size of different individuals are different, and HRTF is closely related to the physiological structure and size, so it is a physical quantity with obvious personal characteristics. [0003] The filter...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): H04S1/00G06K9/62
CPCH04S1/002H04S2420/01G06F18/2136G06F18/2135
Inventor 曾向阳路东东王海涛周治宇马慧颖晋安其
Owner NORTHWESTERN POLYTECHNICAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products