Personalized head-related transfer function modeling method based on deep neural network

A deep neural network, head-related transmission technology, applied in the field of head-related transfer function, personalized head-related transfer function modeling, can solve the problem of not being able to predict HRTF and so on

Active Publication Date: 2018-09-28
PEKING UNIV
View PDF5 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method requires a separate HRTF mapping for each spatial direction, and cannot predict the HRTF of the unsampled spatial directions in the HRTF database.

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
  • Personalized head-related transfer function modeling method based on deep neural network
  • Personalized head-related transfer function modeling method based on deep neural network
  • Personalized head-related transfer function modeling method based on deep neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The preferred embodiments of the present invention will be described in more detail below with reference to the accompanying drawings of the present invention.

[0045] The specific implementation steps of the method of the invention include data preprocessing, spatial principal component analysis and neural network modeling. The present invention adopts CIPIC database, which contains HRIR data of D=1250 directions, and the sampling rate is 44.1kHz. The specific implementation process of each step is as follows:

[0046] 1. Data preprocessing

[0047] To preprocess the original HRIR data in the CIPIC library, the specific steps are as follows:

[0048] First, transform the HRIR data into the frequency domain. Perform Fourier transform on the HRIR in the CIPIC library to obtain the HRTF in the frequency domain. The HRTF in the frequency domain is the horizontal angle θ, the elevation angle A function of the frequenc...

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 a personalized head-related transfer function modeling method based on a deep neural network. According to the method, HRTF data is decomposed on the basis of spatial principalcomponent analysis, and spatial principal components, spatial principal component coefficients and an average spatial function are all modeled by a neural network, wherein the spatial principal components and the average spatial function are only related to a spatial direction, and the spatial principal component coefficients are functions of a frequency and personalized characteristic parameters. The deep neural network is utilized to model the spatial principal components, the average spatial function and an interaural time difference, spatial direction information such as a horizontal angle and an elevation angle is introduced into a network input layer, and the neural network is utilized to model spatial principal component coefficients on the basis of human body measurement parameters. On the basis of the above model, a personalized HRTF in any direction in space can be obtained according to a small number of human body measurement parameters of a subject.

Description

technical field [0001] The invention belongs to the technical field of signal processing, relates to a head-related transfer function, in particular to a modeling method of a personalized head-related transfer function based on spatial principal component analysis. Background technique [0002] Hearing plays a very important role in human life. It senses the sound of the surrounding environment to make corresponding judgments and decisions. In addition to subjective attributes such as the intensity, pitch, and timbre of sound, human hearing can also make judgments about the direction and distance of the sound source. After the sound wave from the sound source reaches the human ears, it is finally perceived by the scattering and reflection of the listener's head, auricle, torso and other human body structures. The study of human's ability to perceive the spatial characteristics of sound has always been a research topic in acoustics and auditory psychology, and has very impo...

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): G06K9/00G06K9/62G06N3/04
CPCG06N3/04G06F2218/12G06F18/2135G06F18/2411
Inventor 曲天书吴玺宏张梦帆
Owner PEKING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap