Publications by authors named "Venkata Pavan Kumar Miriyala"
Spiking neural networks (SNNs), which are a form of neuromorphic, brain-inspired AI, have the potential to be a power-efficient alternative to artificial neural networks (ANNs). Spikes that occur in SNN systems, also known as activations, tend to be extremely sparse, and low in number. This minimizes the number of data accesses typically needed for processing.
View Article and Find Full Text PDFArticle Synopsis
- Spiking neural networks (SNNs) leverage spike patterns for information processing, offering a biological approach that is energy-efficient for neuromorphic systems, but training them is challenging due to the inapplicability of traditional backpropagation algorithms.
- The study introduces a novel spike-timing-dependent backpropagation (STDBP) method and a new rectified linear postsynaptic potential function (ReL-PSP), allowing DeepSNNs to learn effectively by focusing on the timing of spikes.
- Experimental results indicate that DeepSNNs using the STDBP algorithm achieve high classification accuracy, while neuromorphic hardware implementing this model demonstrates ultra-low power consumption (0.751 mW) and quick image classification