Original title: Asynchronous Bioplausible Neuron for Spiking Neural Networks for Event-Based Vision
Authors: Sanket Kachole, Hussain Sajwani, Fariborz Baghaei Naeini, Dimitrios Makris, Yahya Zweiri
The article explores Spiking Neural Networks (SNNs) for computer vision, mimicking how our brains process visual information. SNNs promise efficient data processing but struggle to stay balanced amid diverse and unpredictable inputs. To tackle this, they introduce the Asynchronous Bioplausible Neuron (ABN), a dynamic firing mechanism that adapts to input variations. Testing it across different datasets shows that ABN improves image classification and segmentation while keeping neural balance and using energy more efficiently. Essentially, this innovation helps SNNs handle various visual inputs better, leading to improved performance and more energy-efficient processing—a step forward in creating brain-inspired systems for computer vision.
Original article: https://arxiv.org/abs/2311.11853