Hardware design of LIF with Latency neuron model with memristive STDP synapses

Acciarito, Simone; Cardarilli, G.C.; Cristini, A.; Nunzio, Luca Di; Fazzolari, Rocco; Khanal, Gaurav Mani; Re, M.; Susi, Gianluca

doi:10.1016/j.vlsi.2017.05.006

Cited by 33 publications

(11 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The LIFL (Cardarilli et al, 2013 ; Susi, 2015 ; Acciarito et al, 2017 ) is a neuron model similar to the classical Leaky Integrate-and-Fire (LIF), but characterized by the presence of the spike latency neurocomputational feature (Izhikevich, 2004 ; Cristini et al, 2015 ; Susi et al, 2016 ). The spike latency is a potential-dependent delay time between the overcoming of the “threshold” and the actual spike generation (Izhikevich, 2004 ; Salerno et al, 2011 ).…”

Section: Methodsmentioning

confidence: 99%

“…τ + and τ − are positive time constants for long-term potentiation (LTP, Equation 7a) and long-term depression (LTD, Equation 7c), respectively; A + and A − (positive and negative values, respectively) are the maximum amplitudes of potentiation and depression which are chosen as absolute changes, as in other works (e.g., Acciarito et al, 2017 ). Then, a weight is increased or decreased depending on the pulse order ( pre- before post- , or post- before pre- , respectively; see Figure 2 ).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A Neuro-Inspired System for Online Learning and Recognition of Parallel Spike Trains, Based on Spike Latency, and Heterosynaptic STDP

et al. 2018

Self Cite

View full text Add to dashboard Cite

Humans perform remarkably well in many cognitive tasks including pattern recognition. However, the neuronal mechanisms underlying this process are not well understood. Nevertheless, artificial neural networks, inspired in brain circuits, have been designed and used to tackle spatio-temporal pattern recognition tasks. In this paper we present a multi-neuronal spike pattern detection structure able to autonomously implement online learning and recognition of parallel spike sequences (i.e., sequences of pulses belonging to different neurons/neural ensembles). The operating principle of this structure is based on two spiking/synaptic neurocomputational characteristics: spike latency, which enables neurons to fire spikes with a certain delay and heterosynaptic plasticity, which allows the own regulation of synaptic weights. From the perspective of the information representation, the structure allows mapping a spatio-temporal stimulus into a multi-dimensional, temporal, feature space. In this space, the parameter coordinate and the time at which a neuron fires represent one specific feature. In this sense, each feature can be considered to span a single temporal axis. We applied our proposed scheme to experimental data obtained from a motor-inhibitory cognitive task. The results show that out method exhibits similar performance compared with other classification methods, indicating the effectiveness of our approach. In addition, its simplicity and low computational cost suggest a large scale implementation for real time recognition applications in several areas, such as brain computer interface, personal biometrics authentication, or early detection of diseases.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

A Neuro-Inspired System for Online Learning and Recognition of Parallel Spike Trains, Based on Spike Latency, and Heterosynaptic STDP

et al. 2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…In fact, they allow modulating the amplitude of transmembrane ionic currents I ATDe and I VNRe which represent the couplings between the SA ad HP pacemakers and AT and VN muscles. [20,30] To control the time interval between waves (PR, RR, STinterval) we used the coupling coefficients. They provide proper synchronization behavior of pacemakers in a wide range of heart rhythms.…”

Section: Parameters Obtained Through the Neural Networkmentioning

confidence: 99%

Relationship between Mathematical Parameters of Modified Van der Pol Oscillator Model and ECG Morphological Features

Silvestri¹,

Acciarito²,

Khanal³

2019

Int. J. Adv. Sci. Eng. Inf. Technol.

Self Cite

View full text Add to dashboard Cite

The mathematical model describes the electrical and mechanical activity of the cardiac conduction system thought set of differential equations. By changing the value of parameters included in these equations, it is possible to change the amplitude and the period of ECG waves. Although this model is a powerful tool for modeling the electrical activity of the heart, its use is often limited to those familiar with the differential equations that describe the system. The purpose of this work is to provide a system that allows generating an ECG signal using Ryzhii model without knowing the details of differential equations. First, we provide the relationships between the ECG wave features and the model parameters; then we generalize them through fitting neural networks. Finally, putting in series fitting neural network and heart model, we provide a system that allows generating a synthetic signal by setting as input only the morphological ECG feature. We computed numerical simulation in Simulink environment and implemented the fitting neural networks in Matlab. Results show that non-linear trends characterize the correlation functions between ECG morphological features and model parameters and that the fitting neural networks can generalized this trend by providing the model parameters given in input the respectively ECG feature.

show abstract

“…The SNN that we implemented is based on the biologically plausible leaky integrate and fire with latency (LIFL) neuron model [18]- [22]. Firstly, we implemented the system equations in Matlab®; then, we have realized the schematic of such system in PSpice®, exploiting a circuit previously developed by our group [23], [24]. Finally, the model has been validated to verify whether 1) it observes the fundamental properties of the dynamical relaying mechanisms described in computational neuroscience studies, and 2) if the circuit implementation presents the same behaviour of the mathematical model.…”

Section: Introductionmentioning

confidence: 99%

Towards Neuro-Inspired Electronic Oscillators Based on The Dynamical Relaying Mechanism

Susi

Acciarito

Pascual

et al. 2019

Int. J. Adv. Sci. Eng. Inf. Technol.

Self Cite

View full text Add to dashboard Cite

Electronic oscillators are used for the generation of both continuous and discrete signals, playing a fundamental role in today's electronics. In both contexts, these systems require stringent performances such as spectral purity, low phase noise, frequency and temperature stability. In state of the art oscillators the preservation of some of these aspects is jeopardized by specific critical issues, e.g., the sensitivity to load capacitance or the component aging over time. This leaves room for the search of new technologies for their realization. On the other hand, in the last decade electronics has been influenced by a growing number of neuro-inspired mechanisms, which allowed for alternative techniques aimed at solving some classical critical issues. In this paper we present an exploratory study for the development of electronic oscillators based on the neuro-inspired mechanism dynamical relaying, which relies on a structure composed of three delay coupled units (as neurons or even neuron populations) able to resonate and self-organise to generate and maintain a given rhythm with great reliability over a considerable parameter range, showing robustness to noise. We used the recent leaky integrated and fire with latency (LIFL) as neuron model. We have initially developed the mathematical model of the neuro-inspired oscillator, and implemented it using Matlab®; then, we have realized the schematic of such system in PSpice®. Finally, the model has been validated to verify whether it observes the fundamental properties of the dynamical relaying mechanisms described in computational neuroscience studies, and if the circuit implementation presents the same behaviour of the mathematical model. Validation results suggest that the dynamical relaying mechanism can be proficuously taken in consideration as alternative strategy for the design of electronic oscillators.Keywords-LIFL neuron model; electronic oscillators; dynamical relaying; spiking neural networks.

show abstract

Hardware design of LIF with Latency neuron model with memristive STDP synapses

Cited by 33 publications

References 62 publications

A Neuro-Inspired System for Online Learning and Recognition of Parallel Spike Trains, Based on Spike Latency, and Heterosynaptic STDP

A Neuro-Inspired System for Online Learning and Recognition of Parallel Spike Trains, Based on Spike Latency, and Heterosynaptic STDP

Relationship between Mathematical Parameters of Modified Van der Pol Oscillator Model and ECG Morphological Features

Towards Neuro-Inspired Electronic Oscillators Based on The Dynamical Relaying Mechanism

Contact Info

Product

Resources

About