International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 11 | Nov 2025
p-ISSN: 2395-0072
www.irjet.net
SPIKING NEURAL NETWORKS FOR REAL TIME NEOROMORPHIC COMPUTING: A BIO-INSPIRED APPROACH TO EFFICIENT AI D. Keerthi1, S. Shanthini2 1PG Student, Department of Computer Applications, Jaya College Of Arts & Science, Chennai. 2 Assistant Professor, Department. of Computer Application, Jaya College Of Arts & Science, Chennai.
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The simulation that spiked the communication of
first step. In this paper, we discuss hardware implementations that support energy-efficient computation and use a hybrid approach of gradient descent and spiketiming-dependent plasticity (STDP) to improve the learning efficiency and real-time adaptability of SNNs.
biological neurons, Spiking Neural Networks (SNNs) provide a bio-inspired framework for real-time and energy-efficient neuromorphic computing. SNNs process information through discrete temporal spikes, as opposed to conventional artificial neural networks, which rely on continuous activations. This allows for event-driven computation with a significantly lower power consumption. In order to achieve both biological plausibility and high computational performance, this paper proposes a novel method for designing efficient SNN architectures that combines gradient-based optimization with spike-timing-dependent plasticity (STDP).The suggested approach improves learning capabilities in real-time contexts like robotics, autonomous systems, and edge AI applications by introducing adaptive neuron models with dynamic threshold regulation. Furthermore, with an emphasis on memoryefficient architectures and parallel processing, we investigate neuromorphic hardware integration for low-latency inference. When compared to traditional deep learning models, experimental evaluation on common neuromorphic datasets shows notable gains in energy efficiency, lower computational overhead, and competitive accuracy. By fusing cutting-edge hardware acceleration with biologically inspired learning mechanisms, this research contributes to the creation of scalable, low-power AI systems that can function independently in dynamic, real-world scenarios. The results offer a way forward for neuromorphic intelligence of the next generation, which will be more adaptive and computationally efficient.
1.2 Literature Review [1]. Early and contemporary work emphasizes that SNNs are not only more biologically plausible but also suitable for low-power, latency-sensitive applications when paired with appropriate learning rules and neuromorphic hardware. PMC [2]. STDP alone can be insufficient for complex supervised tasks; as a result, hybrid approaches combining local plasticity with global gradient-based fine-tuning (e.g., surrogate gradients) have been proposed to achieve higher task accuracy while preserving event-driven benefits. PMC+1 Simulation frameworks and toolkits have enabled rapid SNN experimentation and reproducible research. [3]. For machine-learning-oriented SNN development (and integration with gradient-based methods), libraries built on deep learning backends such as BindsNET (PyTorchbased) provide support for hybrid training, large-scale simulations, and easier deployment workflows. These platforms have accelerated progress in bridging biologically inspired mechanisms with practical ML pipelines. PMC+1
[4]. Nonetheless, recent comparative analyses emphasize that energy advantages are context dependent and hinge on architecture, model sparsity, data modality, and mapping strategy; careful benchmarking against optimized ANN implementations is therefore essential when claiming energy benefits.
Key Words: Spiking Neural Network (SNNS), Neuro morphic-Computing, Spike-Timing-Dependent Plasticity(STDP).
1. INTRODUCTION Growing interest in bio-inspired alternatives is a result of conventional Artificial Neural Networks' (ANNs) shortcomings in terms of energy efficiency and real-time responsiveness. Spiking neural networks (SNNs) use discrete spikes to transmit information, simulating the actions of biological neurons. Low-power, real-time data processing is made possible by this event-driven method, which makes it perfect for edge AI and robotics applications. In order to achieve brain-like computation and allow AI systems to function autonomously in dynamic environments, SNN integration with neuromorphic hardware is a promising
© 2025, IRJET
|
Impact Factor value: 8.315
2. COMMON TASKS
|
The creation of energy-efficient SNN architectures for real-time neuromorphic computing is one of the main goals of this study. To enhance learning capability by combining gradient-based optimization with STDP. To present models of adaptive neurons with dynamic threshold control.
ISO 9001:2008 Certified Journal
|
Page 901