Method for personalizing head-related transfer function based on pre-training model
A head-related transmission and pre-training technology, applied in the field of signal processing in the electronics industry, can solve problems such as large errors and long time consumption, and achieve the effects of improved performance, simple use, and high application value
Pending Publication Date: 2021-12-17
CHINA UNIVERSITY OF POLITICAL SCIENCE AND LAW
View PDF0 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
One is a perception-based method, which determines parameters through audiometry experiments. This method needs to measure a large-scale database in advance for matching, and obtain the most suitable HRTF for the target object, which takes a long time
The full space estimation method of HRTF from small data measurement set is also another method of HRTF personalization. However, most of the existing methods only obtain the coefficients of the linear prediction model from the small data test set, and then expand to the full space. large error in
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0037] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail below in conjunction with specific embodiments and with reference to the accompanying drawings.
[0038] It should be noted that, in the accompanying drawings or descriptions in the specification, similar or identical parts all use the same figure number. And in the accompanying drawings, it is marked for simplicity or convenience. Furthermore, implementations not shown or described in the accompanying drawings are forms known to those of ordinary skill in the art. Additionally, while illustrations of parameters containing particular values may be provided herein, it should be understood that the parameters need not be exactly equal to the corresponding values, but rather may approximate the corresponding values within acceptable error margins or design constraints.
[0039] The present invention provides a per...
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
Login to View More Abstract
The invention discloses a method for personalizing a head-related transfer function based on a pre-training model, and particularly relates to the technical field of electronic industry signal processing. The method uses an HRTF feature preprocessing module, a multi-database pre-training module, a physiological feature preprocessing module, a personalizing model fine tuning module and a human body physiological parameter measurement module. The HRTF feature preprocessing module is used for preprocessing a header-related transfer function; and the multi-database pre-training module is used for generating a position-related and user-independent pre-training model. According to the method, the pre-training model is obtained by adopting a plurality of public databases, information related to positioning and personalizing in available data is fully utilized, and more accurate estimation of the personalized head-related transfer function is obtained. In practical application, only the pre-training model needs to be finely adjusted according to the personalized physiological features, the complexity is low, the use is simple, the personalized head-related transfer function can be quickly generated for any target object, and the practical value is relatively high.
Description
technical field [0001] The invention relates to the technical field of signal processing in the electronics industry, in particular to a method for personalizing head-related transfer functions based on pre-trained models. Background technique [0002] The key problem of virtual hearing technology is to restore the same spatial position characteristics as natural hearing. The human auditory process can usually be regarded as a sound source-channel-receiver model, in which the channel includes the diffraction and interference of the sound source through different parts of the human body, and finally reaches the tympanic membrane, which can be regarded as a spatial digital filter, called head-related transmission Function (Head-Related Transfer Function, HRTF). Therefore, HRTF contains all spectral features caused by the interaction between sound waves and body parts, and is highly correlated with human physiological characteristics, such as height, head size, and auricle sha...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/10G06F17/16G06F30/27
CPCG06F17/10G06F17/16G06F30/27
Inventor 戚肖克
Owner CHINA UNIVERSITY OF POLITICAL SCIENCE AND LAW