Spike Neural Input Conversion via Derivative Zero-Point Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing spike neural networks require discrete input signals, while continuous time series data needs conversion for processing.
Innovation Solution
A data processing device and method that converts time series data into discrete data using a differentiator to determine maximum and minimum points, generating spike train signals based on differential values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If continuous time series data is input directly into a spike neural network, then the data can be processed without conversion, but the spike neural network cannot operate because it requires discrete spike signals
Solution Approach 1:
The patent introduces a data processing device as an intermediary component between the continuous data source and the spike neural network. This device converts continuous time series data into discrete spike train signals through differentiation and threshold detection, enabling the neural network to process continuous data while maintaining its requirement for discrete spike inputs.
Solution Approach 2:
The patent transforms the data representation by changing its fundamental parameters: converting from continuous amplitude values to discrete temporal patterns. The conversion process detects zero-crossings of the derivative signal and generates spike trains with specific timing patterns, fundamentally altering the data format to match the neural network's input requirements.
2Reliability
If a data processing device converts time series data into discrete spike signals, then the spike neural network can process the data, but additional conversion steps and devices are required
Solution Approach 1:
The data processing device is divided into distinct functional modules: a differentiation unit that computes the derivative of the input signal, a detection unit that identifies zero-crossing points, and a generation unit that creates spike train outputs. This segmentation allows each component to perform a specific function, making the overall conversion process more manageable and implementable.
Solution Approach 2:
The conversion device uses the inherent properties of the input signal itself to generate the conversion criteria. By differentiating the input signal and detecting its own zero-crossings, the system automatically determines when to generate spikes without requiring external timing information or manual parameter setting, making the conversion process self-regulating.
3Productivity
If the data processing device uses differentiation to convert time series data, then discrete spike signals can be generated, but points where the differential value is zero may be missed
Solution Approach 1:
The system performs preliminary differentiation of the input signal before detecting zero-crossings. By computing the derivative in advance and storing it, the system can systematically search for zero-crossing points without missing any, ensuring that all relevant transition points are captured for spike generation.
Solution Approach 2:
The system implements a feedback mechanism where the detected zero-crossing points are used to generate spike train outputs, which can then be fed back into the neural network. The neural network's responses can further inform subsequent processing, creating a closed-loop system that continuously refines its operation based on actual performance and output requirements.
Data Source
AI summary
Disclosed is a data processing device of a spike neural network. The data processing device according to the present disclosure includes an input terminal that receives time series data, a processor that differentiates the input time series data, determines a point at which a differential value of the time series data is “0”, and generates an output in a discrete form, based on the determination result, and an output terminal that generates a spike train signal, based on the output and outputs the generated spike train signal to external axons.