Voltage-Dependent Synaptic Plasticity (VDSP): Unsupervised probabilistic Hebbian plasticity rule based on neurons membrane potential

TL;DR

This paper introduces VDSP, a novel unsupervised, voltage-dependent synaptic plasticity rule inspired by Hebbian learning, which efficiently updates synapses based on membrane potential, demonstrating high accuracy in digit recognition tasks on neuromorphic hardware.

Contribution

The paper presents VDSP, a new local learning rule that reduces update frequency, depends on presynaptic membrane potential, and is mathematically equivalent to STDP, with validated performance on MNIST.

Findings

Achieves over 85% accuracy on MNIST with 100 output neurons.

Performance improves with larger network sizes, reaching over 90% accuracy.

Better adaptation to input frequency compared to STDP, without hyperparameter tuning.

Abstract

This study proposes voltage-dependent-synaptic plasticity (VDSP), a novel brain-inspired unsupervised local learning rule for the online implementation of Hebb's plasticity mechanism on neuromorphic hardware. The proposed VDSP learning rule updates the synaptic conductance on the spike of the postsynaptic neuron only, which reduces by a factor of two the number of updates with respect to standard spike-timing-dependent plasticity (STDP). This update is dependent on the membrane potential of the presynaptic neuron, which is readily available as part of neuron implementation and hence does not require additional memory for storage. Moreover, the update is also regularized on synaptic weight and prevents explosion or vanishing of weights on repeated stimulation. Rigorous mathematical analysis is performed to draw an equivalence between VDSP and STDP. To validate the system-level performance of VDSP, we train a single-layer spiking neural network (SNN) for the recognition of handwritten digits. We report 85.01 $ \pm $ 0.76% (Mean $ \pm $ S.D.) accuracy for a network of 100 output neurons on the MNIST dataset. The performance improves when scaling the network size (89.93 $ \pm $ 0.41% for 400 output neurons, 90.56 $ \pm $ 0.27 for 500 neurons), which validates the applicability of the proposed learning rule for spatial pattern recognition tasks. Future work will consider more complicated tasks. Interestingly, the learning rule better adapts than STDP to the frequency of input signal and does not require hand-tuning of hyperparameters