Personalized Acoustic Transfer Function Modeling via Statistical Adaptation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for measuring individual acoustic transfer functions are cumbersome, expensive, and difficult to implement for the general public due to the need for specialized equipment and lengthy procedures, making it challenging to achieve high-quality binaural synthesis for spatialized sound reproduction.

Innovation Solution

A method using statistical analysis and psychophysical inverse correlation techniques to model individual acoustic transfer functions, which involves transmitting stimuli and receiving responses to calculate personalized acoustic transfer functions, reducing the need for extensive equipment and measurement procedures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional measurement methods are used to obtain individual acoustic transfer functions, then measurement precision is improved, but device complexity and ease of operation deteriorate due to requiring anechoic chambers and specialized equipment

Engineering Contradiction:
Improveaccuracy of acoustic transfer function measurementVSAvoidcomplexity of measurement equipment
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses pre-measured acoustic transfer functions from a database (copies from other individuals) instead of requiring direct measurement on each user. A selection of HRTFs is made from the database based on morphological parameters, and these are then adapted to the target user through statistical analysis of their responses, eliminating the need for complex measurement equipment while maintaining reasonable accuracy

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent transforms the measurement problem into a parameter estimation problem. Instead of directly measuring acoustic transfer functions, the system estimates user-specific parameters (morphological features, head dimensions, ear canal length) and uses these to select and adapt HRTFs from a database, changing the approach from physical measurement to parameter-based selection and statistical adaptation

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If conventional measurement methods are used to obtain individual acoustic transfer functions, then measurement precision is improved, but loss of time increases due to lengthy measurement procedures

Engineering Contradiction:
Improveaccuracy of acoustic transfer function measurementVSAvoidduration of measurement procedure
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs measurements on a panel of individuals in advance, storing their acoustic transfer functions and morphological parameters in a database. When a new user needs HRTFs, the system retrieves pre-computed data from the database and adapts it statistically, eliminating the need for time-consuming on-site measurements while maintaining individualization through response-based adaptation

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system allows users to self-test and self-select from the database using simple response indicators (e.g., head movements, button presses) rather than requiring professional measurement technicians. The statistical adaptation process automatically adjusts the selected HRTFs based on user responses, enabling users to obtain personalized HRTFs without professional assistance

Inventive Principle:
Principle #25Self-service

3Ease of operation

If database selection methods are used to choose acoustic transfer functions, then ease of operation is improved, but measurement precision deteriorates due to lack of individualization

Engineering Contradiction:
Improvesimplicity of HRTF selection processVSAvoidaccuracy of acoustic transfer function matching
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where the user's responses to test stimuli (head movements, localization accuracy) are used to statistically adapt the selected HRTFs. This feedback loop allows the system to refine the match between the user's actual acoustic characteristics and the database HRTFs, improving precision while maintaining the simplicity of database selection

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent makes the HRTF selection dynamic rather than static. Instead of simply selecting a fixed HRTF from the database, the system continuously adapts the selected HRTFs based on real-time user responses, allowing the acoustic transfer functions to dynamically adjust to the user's actual morphology and preferences, thereby improving precision while keeping the interface simple

Inventive Principle:
Principle #15Dynamics

4Measurement precision

If extensive measurements are performed to cover all spatial directions, then measurement precision is improved, but productivity decreases due to the large number of measurements required

Engineering Contradiction:
Improvecompleteness of spatial coverageVSAvoidefficiency of HRTF acquisition
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the HRTF acquisition process into two parts: (1) comprehensive measurements on a panel of individuals performed once in advance, and (2) rapid selection and adaptation for each new user. This segmentation allows the expensive, time-consuming measurements to be performed only once for multiple users, dramatically improving productivity while maintaining precision through the use of panel data

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates a universal database of HRTFs measured on a diverse panel of individuals that can serve multiple users. This universal resource is then customized for each user through statistical adaptation, allowing the system to achieve high precision for many users without repeating the expensive measurement process, thereby improving productivity while maintaining completeness of spatial coverage

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3484185B1Modelling of a set of acoustic transfer functions suitable for an individual, three-dimensional sound card and system for three-dimensional sound reproduction
Publication Date: 2024.07.31 ORANGE SA
  • EP3484185B1 patent drawingFigure 1
  • EP3484185B1 patent drawingFigure 2~3

AI summary

The invention relates to the modeling of individual acoustic transfer functions, pertaining to an individual's hearing in three-dimensional space. An object of the invention is a method for modeling sets of acoustic transfer functions specific to an individual along a multiplicity of spatial directions, wherein a set of acoustic transfer functions specific to an individual in a given direction of the multiplicity of directions is determined based on the result of a statistical analysis of several distinct stimuli emitted towards the individual, a stimulus being a function of at least one set of predetermined acoustic transfer functions associated with the given direction, and of responses received from the individual to each emitted stimulus.Thus, the invention is more reliable and robust than a simple selection of a set of acoustic transfer functions in a database and overcomes the disadvantage of the critical acquisition of the 3D mesh of the individual morphology used by classical numerical modeling.