Neuromorphic Sound Source Localization Circuit

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current digital audio signal processing methods for locating sound sources, such as those used in robots and security cameras, are inefficient in terms of power consumption and processing speed, and do not effectively utilize neuromorphic principles to enhance sound source localization.

Innovation Solution

A neuromorphic signal processing device utilizing a plurality of neuron circuits, including a detector, multiplexor, and integrator, which outputs a multiplexed spiking signal corresponding to a predetermined time difference by leveraging delay and coincidence detection neuron circuits, and an integration neuron circuit with excitatory and inhibitory synapses to enhance sound source localization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If digital audio signal processing methods are used for sound source localization, then measurement precision can be achieved, but power consumption increases and processing speed decreases

Engineering Contradiction:
Improvesound source localization precisionVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent replaces digital audio signal processing systems with a neuromorphic signal processing system that mimics biological auditory processing. The neuromorphic device uses spiking neural networks to perform sound source localization, substituting conventional digital signal processing mechanisms with bio-inspired neural circuitry that operates more efficiently in terms of power consumption while maintaining localization precision.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the operational parameters from digital continuous signal processing to spiking temporal code processing. By transforming the representation of auditory signals from amplitude-based digital waves to time-based spike trains, the system achieves efficient power consumption characteristics while preserving measurement precision for sound source localization.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If digital audio signal processing methods are used for sound source localization, then measurement precision can be achieved, but processing speed decreases

Engineering Contradiction:
Improvesound source localization precisionVSAvoidprocessing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces sequential digital signal processing operations with parallel neuromorphic processing. The spiking neural network architecture enables simultaneous processing of multiple frequency bands and temporal differences, dramatically increasing processing speed while maintaining the precision required for accurate sound source localization.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent implements preliminary temporal encoding of auditory signals into spiking patterns that pre-organize information for rapid processing. By converting analog auditory inputs into temporally-coded spike trains before processing, the system prepares data in a format that enables fast parallel operations in the neuromorphic network, improving overall processing speed without sacrificing localization precision.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If conventional signal processing components are used, then sound source localization can be performed, but device complexity increases

Engineering Contradiction:
Improvesound source localization capabilityVSAvoidnumber of components
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple signal processing functions into integrated neuromorphic circuit blocks. The spiking neural network architecture combines temporal difference analysis, frequency band processing, and sound source localization into a unified neural system, reducing the number of discrete components while maintaining comprehensive localization capability.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent creates a universal neuromorphic processing platform that performs multiple functions through a single spiking neural network architecture. The same neural circuitry handles temporal difference detection, frequency analysis, and localization computation, eliminating the need for separate dedicated components for each function and thereby reducing overall device complexity.

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

Data Source

PatentEP2672432B1Neuromorphic signal processing device and method for locating sound source using a plurality of neuron circuits
Publication Date: 2021.08.04 SAMSUNG ELECTRONICS CO LTD
  • EP2672432B1 patent drawingFigure 1
  • EP2672432B1 patent drawingFigure 2
  • EP2672432B1 patent drawingFigure 3

AI summary

Provided is a neuromorphic signal processing device for locating a sound source using a plurality of neuron circuits, the neuromorphic signal processing device including a detector configured to output a detected spiking signal using a detection neuron circuit corresponding to a predetermined time difference, in response to a first signal and a second signal containing an identical input spiking signal with respect to the predetermined time difference, for each of a plurality of predetermined frequency bands, a multiplexor configured to output a multiplexed spiking signal corresponding to the predetermined time difference based on a plurality of the detected spiking signals output from a plurality of neuron circuits corresponding to the plurality of frequency bands, and an integrator configured to output an integrated spiking signal corresponding to the predetermined time difference, based on a plurality of the multiplexed spiking signals corresponding to a plurality of predetermined time differences.