Neuron Element Group Information Recognition via Synapse Circuits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing neuron element-based systems for recognizing group information are prone to responding to non-matching inputs due to increased capacitance across multiple electrodes, leading to false activation.
Innovation Solution
Incorporating a synapse determination circuit with excitatory and inhibitory synapses that apply specific voltage levels to a multi-input gate electrode, allowing the neuron element to learn and respond only to identical group information by accumulating electric charges in a ferroelectric film, and using a dielectric film that does not accumulate charges based on voltage application frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If voltage is applied to multiple electrodes in the multi-input gate electrode to increase capacitance for learning group information, then the neuron element becomes more responsive to stored patterns, but the field-effect transistor may respond and output a signal even when voltage is applied to electrodes other than those corresponding to the stored group information, leading to false activation
Solution Approach 1:
The gate electrode is divided into multiple independent electrodes (first gate electrode, second gate electrode, third gate electrode), each corresponding to specific group information. This segmentation allows selective voltage application to individual electrodes, enabling the neuron element to distinguish between different input patterns and respond only to the stored pattern, thereby preventing false activation while maintaining response accuracy.
2Adaptability or versatility
If a ferroelectric film is used to accumulate electric charges based on voltage application frequency for learning, then the neuron element can store and recognize group information patterns, but the system complexity increases due to the need for precise control of voltage application timing and frequency
Solution Approach 1:
The ferroelectric film is pre-configured with specific polarization characteristics that correspond to the group information to be learned. By applying voltage in a predetermined sequence and frequency pattern during the learning phase, the ferroelectric film accumulates electric charges that encode the group information. This preliminary configuration simplifies the recognition phase, as the neuron element automatically responds to matching inputs without requiring complex real-time control.
3Measurement precision
If the capacitance between the gate electrode and floating gate electrode is increased to enhance learning capability, then the neuron element becomes more sensitive to input patterns, but the threshold voltage requirement becomes harder to meet, requiring higher voltage application
Solution Approach 1:
The capacitance enhancement is applied locally at specific interfaces where the ferroelectric film is positioned between the gate electrode and floating gate electrode, rather than uniformly across the entire device. This localized capacitance increase improves pattern detection sensitivity at the critical sensing interface while minimizing the overall voltage requirement and energy consumption of the field-effect transistor.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively learns and recognizes specific group information patterns, reducing false activations and ensuring accurate response only to identical inputs by controlling voltage levels and synapse types.
Implementation Method 1
a first ferroelectric film formed over the floating gate electrode and accumulating electric charges according to the number of times of voltage application
Implementation Method 2
using a dielectric film that does not accumulate charges based on voltage application frequency
Implementation Method 3
recognizing whether another pattern subsequently inputted is identical with the pattern that is stored; and when another pattern is identical, it outputs a signal in response thereto
Data Source
AI summary
A plurality of synapse determination circuits are provided on a one-to-one basis for a plurality of gate electrodes of a multi-input gate electrode in a neuron element. With respect to first image regions where “1” is repeatedly inputted in correspondence with group information, the synapse determination circuits corresponding to the first image regions are excitatory synapses. With respect to second image regions where “0” is repeatedly inputted in correspondence with the group information, the synapse determination circuits corresponding to the second image regions are inhibitory synapses.