Individualized HRTF Generation via Acoustic Feature-Based Notch Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audio systems fail to provide personalized head-related transfer functions (HRTFs) that accurately simulate sound source direction and localization, as they do not account for individual anatomical differences, leading to a non-immersive audio experience.
Innovation Solution
A system that generates individualized HRTFs by applying customized filters to a template HRTF based on user-specific acoustic features data, using machine learning to determine notch parameters, which are then applied to audio data to create spatialized audio content that appears to originate from specific directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a template HRTF is used for all users, then the device complexity is reduced, but the audio localization accuracy deteriorates because individual anatomical differences are not accounted for
Solution Approach 1:
The patent applies local quality by customizing specific frequency notches in the HRTF based on individual user anatomy (ear shape, head size) while keeping the overall HRTF structure standardized. This allows selective personalization of acoustic characteristics without redesigning the entire HRTF system, thus maintaining low complexity while improving localization accuracy for specific frequency ranges.
Solution Approach 2:
The patent changes parameters of the HRTF by adjusting notch frequency, depth, and bandwidth based on user-specific anatomical measurements. This parameter adjustment approach allows the system to adapt to individual users without requiring completely different HRTF sets, resolving the contradiction between simplicity and accuracy.
2Measurement precision
If individualized HRTFs are generated for each user, then the audio localization accuracy is improved, but the device complexity increases due to personalized processing requirements
Solution Approach 1:
The system generates individualized HRTFs by modifying parameters (notch frequency, depth, bandwidth) of a template HRTF based on user anatomical data. This parameter-based approach avoids the complexity of creating entirely new HRTFs for each user while still achieving personalized audio localization.
Solution Approach 2:
The patent uses a template HRTF as a base model that is then customized for individual users. This copying approach allows the system to leverage a pre-designed, optimized HRTF structure while adding only the necessary personalized elements, thus minimizing the increase in system complexity.
3Measurement precision
If acoustic features data collection is performed for each user, then the HRTF individualization accuracy is improved, but the ease of operation deteriorates due to additional setup requirements
Solution Approach 1:
The system automatically captures acoustic features data using the headset's built-in microphones and processors without requiring external equipment or complex user actions. The user simply wears the headset, and the system self-automatically performs measurements and generates personalized HRTFs, maintaining ease of operation while achieving accurate individualization.
Data Source
AI summary
A system for generating individualized HRTFs that are customized to a user of a headset. The system includes a server and an audio system. The server determines the individualized HRTFs based in part on acoustic features data (e.g., image data, anthropometric features, etc.) of the user and a template HRTF. The server provides the individualized HRTFs to the audio system. The audio system presents spatialized audio content to the user using the individualized HRTFs.