Spiking Neuron Network Contrast Enhancement via Inhibitory Signals

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing approaches for detecting salient features in visual inputs, particularly in the presence of variable brightness and color, are inadequate in terms of temporal and spatial response, and spiking neuron networks used for visual attention are often overly complex and slow to adapt to changing conditions.

Innovation Solution

A spiking neuron network apparatus and method that encodes visual input using spike latency, where excitatory signals generate inhibitory signals to suppress responses to non-salient features, allowing for faster detection of salient features by encoding spectral illuminance and other attributes into pulse latency and using inhibitory units to prevent network responses to minor features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If spiking neuron networks are used to encode visual information for visual attention, then the network can process sensory input, but the system becomes overly complex and responds slowly to changing conditions

Engineering Contradiction:

Improveresponse speed to changing conditionsVSAvoidnetwork complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The spiking neuron network is divided into distinct functional segments: excitatory neuron units that generate spikes in response to sensory input, inhibitory neuron units that generate inhibition signals, and sensory input units that receive external stimuli. This segmentation allows each component to perform its specific function efficiently, reducing overall system complexity while maintaining adaptability to changing visual conditions

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The network implements dynamic adaptability through latency-based encoding where the timing of excitatory spikes relative to inhibition signals encodes visual feature salience. The system dynamically adjusts its response based on the temporal relationship between excitatory and inhibitory signals, allowing fast response to changing conditions without increasing structural complexity

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If existing approaches are used to detect salient features in visual input with variable brightness and color, then the system can process visual information, but the temporal and spatial response is inadequate

Engineering Contradiction:

Improvesalient feature detection accuracyVSAvoidtemporal response delay

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary encoding of visual features into latency-coded excitatory signals before salient feature detection is required. By pre-processing sensory input through the excitatory-inhibitory interaction framework and encoding spectral illuminance and spatial attributes into pulse latency, the system prepares the neural representation in advance, enabling faster and more precise detection of salient features when needed

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The inhibitory neuron units receive excitatory spikes as input and generate inhibition signals that feed back to suppress non-salient features. This feedback mechanism allows the system to dynamically identify and suppress non-salient stimuli based on temporal patterns, improving both the precision of salient feature detection and the temporal response by eliminating unnecessary processing of non-salient information

Inventive Principle:

Principle #23Feedback

3Productivity

If inhibitory signals are generated to suppress non-salient features, then salient feature detection is enabled, but the system must process and differentiate multiple signal types

Engineering Contradiction:

Improvesalient feature detection efficiencyVSAvoidsignal processing complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The inhibitory neuron units serve as intermediary elements that translate excitatory spike patterns into inhibition signals. These inhibitory signals act as mediators that suppress non-salient features in the sensory input units, enabling efficient salient feature detection. The intermediary inhibitory layer simplifies the overall processing by providing a clear mechanism for feature suppression without requiring complex direct interactions between all network components

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS9070039B2Temporal winner takes all spiking neuron network sensory processing apparatus and methods

Publication Date: 2015.06.30 BRAIN CORP

AI summary

Apparatus and methods for contrast enhancement and feature identification. In one implementation, an image processing apparatus utilizes latency coding and a spiking neuron network to encode image brightness into spike latency. The spike latency is compared to a saliency window in order to detect early responding neurons. Salient features of the image are associated with the early responding neurons. An inhibitory neuron receives salient feature indication and provides inhibitory signal to the other neurons within an area of influence of the inhibitory neuron. The inhibition signal reduces probability of responses by the other neurons to stimulus that is proximate to the feature thereby increasing contrast within the encoded data. The contrast enhancement may facilitate feature identification within the image. Feature detection may be used for example for image compression, background removal and content distribution.