This paper explores the object classification performance of spiking neural networks (SNNs) using the temporal spike-based backpropagation technique on the Field Programmable Gate Array (FPGA) platform. The FPGA board is specially designed to host the spiking neural networks for artificial intelligence tasks such as object classification, object detection, and segmentation. The publicly available classification datasets such as MNIST, CIFAR10 were employed to examine the performance of the SNNs on the FPGA platform. Similarly, the latest temporal spike-based backpropagation technique was chosen to investigate the neuromorphic ability of the low-cost FPGA board in processing SNNs for object classification tasks. The main purpose of this research proceeding is to facilitate the neuromorphic research community with the information regarding (i). the exploitation of the low-cost FPGA design for neuromorphic image processing and artificial intelligence (AI) tasks; (ii). cross-validating temporal spike-based backpropagation trained SNNs on FPGA alongside PC; (iii). assessing the performance stability and industrial choices of low-cost FPGAs for object classification tasks and related issues. The evaluation metrics such as classification accuracy, mean average precision, and processing time were utilized to assess the performance of the SNN model on FPGA alongside PC. This study will be used as an informative report for the researchers working towards perfecting the neuromorphic hardware for processing SNNs in imminent studies. 

Performance Analysis of Spiking Neural Network using Temporal Spike-based Backpropagation on Field Programmable Gate Array (FPGA) platform

Sleeping posture recognition plays a vital role in various clinical applications. Many studies show that pressure sensor-based solutions work well for assessing in-bed positions. In recent years, Neuromorphic Computing has attracted many researchers' attention due to its advantage of energy efficiency. Surprisingly, the applications of Neuromorphic Computing in sleeping posture classification have been still lacking. This study proposed a novel approach that combines a preprocessing technique and an ensemble model based on a neuromorphic computing architecture called RANC. Experimental results confirm that our proposed method can gain 99.99% and 92.4% accuracy in the Leave-One-Subject-Out (LOSO) validation for 3 and 17 sleeping postures, respectively. This result greatly surpasses the previous SNN-based sleeping posture classification method. 

A novel implementation of sleeping posture classification using RANC ecosystem

Recently, spiking neural networks (SNNs) have been widely studied by researchers due to their biological interpretability and potential application of low power consumption. However, the traditional clock-driven simulators have the problem that the accuracy is limited by the time-step and the lateral inhibition failure. To address this issue, we introduce EvtSNN (Event SNN), a faster SNN event-driven simulator inspired by EDHA (Event-Driven High Accuracy). Two innovations are proposed to accelerate the calculation of event-driven neurons. Firstly, the intermediate results can be reused in population computing without repeated calculations. Secondly, unnecessary peak calculations will be skipped according to a condition. In the MNIST classification task, EvtSNN took 56 s to complete one epoch of unsupervised training and achieved 89.56% accuracy, while EDHA takes 642 s. In the benchmark experiments, the simulation speed of EvtSNN is 2.9–14.0 times that of EDHA under different network scales.

https://www.frontiersin.org/articles/10.3389/fnins.2022.944262/pdf

EvtSNN: Event-driven SNN simulator optimized by population and pre-filtering

Sleeping posture recognition plays a vital role in various clinical applications. Many studies show that pressure sensor-based solutions work well for assessing in-bed positions. In recent years, Neuromorphic Computing has attracted many researchers' attention due to its advantage of energy efficiency. Surprisingly, the applications of Neuromorphic Computing in sleeping posture classification have been still lacking. This study proposed a novel approach that combines a preprocessing technique and an ensemble model based on a neuromorphic computing architecture called RANC. Experimental results confirm that our proposed method can gain 99.99% and 92.4% accuracy in the Leave-One-Subject-Out (LOSO) validation for 3 and 17 sleeping postures, respectively. This result greatly surpasses the previous SNN-based sleeping posture classification method.

This paper explores the object classification performance of spiking neural networks (SNNs) using the temporal spike-based backpropagation technique on the Field Programmable Gate Array (FPGA) platform. The FPGA board is specially designed to host the spiking neural networks for artificial intelligence tasks such as object classification, object detection, and segmentation. The publicly available classification datasets such as MNIST, CIFAR10 were employed to examine the performance of the SNNs on the FPGA platform. Similarly, the latest temporal spike-based backpropagation technique was chosen to investigate the neuromorphic ability of the low-cost FPGA board in processing SNNs for object classification tasks. The main purpose of this research proceeding is to facilitate the neuromorphic research community with the information regarding (i). the exploitation of the low-cost FPGA design for neuromorphic image processing and artificial intelligence (AI) tasks; (ii). cross-validating temporal spike-based backpropagation trained SNNs on FPGA alongside PC; (iii). assessing the performance stability and industrial choices of low-cost FPGAs for object classification tasks and related issues. The evaluation metrics such as classification accuracy, mean average precision, and processing time were utilized to assess the performance of the SNN model on FPGA alongside PC. This study will be used as an informative report for the researchers working towards perfecting the neuromorphic hardware for processing SNNs in imminent studies.

Spiking Neural Network (SNN), the next generation of Neural Network, is supposed to be more energy-saving than the previous generation represented by Convolution Neural Network (CNN). Although CNNs have shown impressive results on various tasks such as natural language processing, image classification, or voice recognition using Graphical Processing Units (GPUs) for training, it is expensive and is not suitable for hardware implementation. The emergence of SNNs is a solution for CNNs in terms of energy consumption. In the dozen types of hardware, Field Programmable Gate Arrays (FPGAs) is a promising approach for SNN implementation on hardware. This paper provides a survey of a number of FGPA-based SNN implementations focused on some aspects such as neuron models, network architecture, training algorithms and applications. The survey provides the reader with a compact and informative insight into recent efforts in this domain.

A review of SNN implementation on FPGA

Image classification is always an interesting problem due to its practical applications in real life. With a capability of self-learning features, modern Convolution Neural Network (CNN) models can achieve high accuracy on large and complex benchmark datasets. However, due to their high computation costs, the CNN models experience energy consumption problems during training and implementation of the hardware which limits their utilisation in mobile and embedded applications. Recently, the Spiking Neural Network (SNN) has been proposed to overcome drawbacks of the CNN models. Like the biological nervous system, the SNN’s neurons communicate with each other by sending spike trains. A neuron is only calculated when a new input spike arrives. As a result, it turns the networks into an energy-saving mode which is suitable for implementation on hardware devices. To avoid the difficulty of the SNN direct training, an indirect training approach is proposed in this work. A proposed CNN model is firstly trained with the RMSprop algorithm then the optimised weights and bias are mapped to the SNN model converted from the proposed CNN model. Experimental results confirm that our model achieves the best accuracy of 93.5% when compared to state-of-the-art SNN approaches on the Fashion- MNIST dataset.

Quoc Trung Pham

Papers

A review of SNN implementation on FPGA

An improved spiking network conversion for image classification