Physio-Sensory Signal Transduction Using Gaussian Envelope Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately translate physiological signals, particularly brain activity, into sensory feedback without requiring parameterization and are limited to synthesizer sounds, lacking flexibility and automation.
Innovation Solution
A method and device that transduce physiological signals into sensory signals using a sum of Gaussian functions to fit raw patterns, allowing for continuous and personalized signal processing without synthesizers, and incorporating slow periodic and constant functions for enhanced sound generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If ADSR language parameterization is used to transcribe physiological signals into sensory signals, then the transcription can be controlled through synthesizer parameters, but the method is limited to synthesizer sounds and requires complex parameterization
Solution Approach 1:
The patent extracts the essential feature of physiological signals (raw patterns) and directly maps them to sensory signal generation without requiring intermediate parameterization through ADSR language. This removes the complex parameterization step while maintaining the core transcription function, thereby reducing device complexity while preserving versatility.
Solution Approach 2:
Instead of parameterizing physiological signals first and then generating sounds (traditional approach), the patent inverts the process by directly using raw physiological patterns to modulate sound generation. This inversion eliminates the need for synthesizer parameterization while expanding sound generation flexibility to include natural sounds and voices.
2Adaptability or versatility
If traditional transcription methods are used, then synthesizer sounds can be generated, but the method cannot produce a wide range of sounds such as voices or natural sounds
Solution Approach 1:
The patent creates a universal transcription method that can generate multiple types of sensory signals (synthesizer sounds, natural sounds, voices) from the same physiological signal input. The system uses raw pattern detection combined with various envelope functions (exponential, power law, Gaussian) to modulate different sound sources, achieving multi-functionality without increasing implementation complexity.
3Measurement precision
If raw patterns are directly used without fitting functions, then the processing is simpler, but the transcription accuracy and continuity are reduced
Solution Approach 1:
The patent applies partial fitting functions (exponential, power law, Gaussian envelopes) to specific portions of raw physiological patterns rather than fully parameterizing the entire signal. This selective application of fitting functions improves transcription accuracy and continuity while avoiding the complexity of complete parameterization, achieving optimal balance between precision and processing simplicity.
Data Source
Figure 1
Figure 2~3
Figure 4A~4B
AI summary
The invention relates to a physio-sensory transduction method comprising: an acquisition (E1) of a physiological signal (10) of an organism comprising a raw sequence (Sr) of at least one raw pattern (100), and a generation (E2) of a sensory signal (30) associated with the raw sequence (Sr). The generation (E2) of the sensory signal (30) comprises a transformation (E21) of the raw sequence (Sr) into a modulated sequence (Sm), said modulated sequence (Sm) comprising at least one modulated pattern (101), wherein each modulated pattern (101) is obtained by fitting the raw pattern (100) with a sum of at least one Gaussian function (fG), said sum of at least one Gaussian function forming an envelope to each raw pattern.