Application Of Spiking Neural Networks Research Articles

Spiking neural networks for biomedical signal analysis.

Artificial intelligence (AI) has had a significant impact on human life because of its pervasiveness across industries and its rapid development. Although AI has achieved superior performance in learning and reasoning, it encounters challenges such as substantial computational demands, privacy concerns, communication delays, and high energy consumption associated with cloud-based models. These limitations have facilitated a paradigm change in on-device AI processing, which offers enhanced privacy, reduced latency, and improved power efficiency through the direct execution of computations on devices. With advancements in neuromorphic systems, spiking neural networks (SNNs), often referred to as the next generation of AI, are currently in focus as on-device AI. These technologies aim to mimic the human brain efficiency and provide promising real-time processing with minimal energy. This study reviewed the application of SNNs in the analysis of biomedical signals (electroencephalograms, electrocardiograms, and electromyograms), and consequently, investigated the distinctive attributes and prospective future paths of SNNs models in the field of biomedical signal analysis.

Low-energy and tunable LIF neuron using SiGe bandgap-engineered resistive switching transistor

We have proposed leaky integrate-and-fire (LIF) neuron having low-energy consumption and tunable functionality without external circuit components. Our LIF neuron has a simple configuration consisting of only three components: one bandgap-engineered resistive switching transistor (BE-RST), one capacitor, and one resistor. Here, the crucial point is that BE-RST with a silicon–germanium heterojunction possesses an amplified hysteric current switching with a low latch-up voltage due to improved hole storage capability and impact ionization coefficient. Therefore, the proposed neuron utilizing BE-RST requires an energy consumption of 0.36 pJ/spike, which is approximately six times lower than 2.08 pJ/spike of pure silicon-RST based neuron. In addition, the spiking properties can be tuned by modulating the leakage rate and threshold through gate bias, which contributes to energy-efficient sparse-activity and high learning accuracy. As a result, our proposed neuron can be a promising candidate for executing various spiking neural network applications.

Spiking-NeRF: Spiking Neural Network for Energy-Efficient Neural Rendering

Artificial Neural Networks (ANNs) have achieved remarkable performance in many artificial intelligence tasks. As the application scenarios become more sophisticated, the computation and energy consumption of ANNs are also constantly increasing, which poses a challenge for deploying ANNs on energy-constrained devices. Spiking Neural Networks (SNNs) provide a promising solution to build energy-efficiency neural networks. However, the current training methods of SNNs cannot output values as precise as ANNs. This limits the applications of SNNs to relatively simple image classification tasks. In this article, we extend the application of SNNs to neural rendering tasks and propose an energy-efficient spiking neural rendering model, called Spiking-NeRF (Spiking Neural Radiance Fields). We first analyze the ANN-to-SNN conversion theory and propose an output scheme for SNNs to obtain the precise scene property values. Then we customize the parameter normalization method for the special network architecture of neural rendering. Furthermore, we present an early termination strategy (ETS) based on the discrete nature of spikes to reduce energy consumption. We evaluate the performance of Spiking-NeRF on both realistic and synthetic scenes. Experimental results show that Spiking-NeRF can achieve comparable rendering performance to ANN-based NeRF with up to \(2.27\times\) energy reduction.

Point-to-Spike Residual Learning for Energy-Efficient 3D Point Cloud Classification

Spiking neural networks (SNNs) have revolutionized neural learning and are making remarkable strides in image analysis and robot control tasks with ultra-low power consumption advantages. Inspired by this success, we investigate the application of spiking neural networks to 3D point cloud processing. We present a point-to-spike residual learning network for point cloud classification, which operates on points with binary spikes rather than floating-point numbers. Specifically, we first design a spatial-aware kernel point spiking neuron to relate spiking generation to point position in 3D space. On this basis, we then design a 3D spiking residual block for effective feature learning based on spike sequences. By stacking the 3D spiking residual blocks, we build the point-to-spike residual classification network, which achieves low computation cost and low accuracy loss on two benchmark datasets, ModelNet40 and ScanObjectNN. Moreover, the classifier strikes a good balance between classification accuracy and biological characteristics, allowing us to explore the deployment of 3D processing to neuromorphic chips for developing energy-efficient 3D robotic perception systems.

RFI detection with spiking neural networks

Abstract Detecting and mitigating radio frequency interference (RFI) is critical for enabling and maximising the scientific output of radio telescopes. The emergence of machine learning (ML) methods capable of handling large datasets has led to their application in radio astronomy, particularly in RFI detection. Spiking neural networks (SNNs), inspired by biological systems, are well suited for processing spatio-temporal data. This study introduces the first exploratory application of SNNs to an astronomical data processing task, specifically RFI detection. We adapt the nearest latent neighbours (NLNs) algorithm and auto-encoder architecture proposed by previous authors to SNN execution by direct ANN2SNN conversion, enabling simplified downstream RFI detection by sampling the naturally varying latent space from the internal spiking neurons. Our subsequent evaluation aims to determine whether SNNs are viable for future RFI detection schemes. We evaluate detection performance with the simulated HERA telescope and hand-labelled LOFAR observation dataset the original authors provided. We additionally evaluate detection performance with a new MeerKAT-inspired simulation dataset that provides a technical challenge for machine-learnt RFI detection methods. This dataset focuses on satellite-based RFI, an increasingly important class of RFI and is an additional contribution. Our SNN approach remains competitive with the original NLN algorithm and AOFlagger in AUROC, AUPRC, and F1-scores for the HERA dataset but exhibits difficulty in the LOFAR and Tabascal datasets. However, our method maintains this accuracy while completely removing the compute and memory-intense latent sampling step found in NLN. This work demonstrates the viability of SNNs as a promising avenue for ML-based RFI detection in radio telescopes by establishing a minimal performance baseline on traditional and nascent satellite-based RFI sources and is the first work to our knowledge to apply SNNs in astronomy.

Trainable Spiking-YOLO for low-latency and high-performance object detection

Spiking neural networks (SNNs) are considered an attractive option for edge-side applications due to their sparse, asynchronous and event-driven characteristics. However, the application of SNNs to object detection tasks faces challenges in achieving good detection accuracy and high detection speed. To overcome the aforementioned challenges, we propose an end-to-end Trainable Spiking-YOLO (Tr-Spiking-YOLO) for low-latency and high-performance object detection. We evaluate our model on not only frame-based PASCAL VOC dataset but also event-based GEN1 Automotive Detection dataset, and investigate the impacts of different decoding methods on detection performance. The experimental results show that our model achieves competitive/better performance in terms of accuracy, latency and energy consumption compared to similar artificial neural network (ANN) and conversion-based SNN object detection model. Furthermore, when deployed on an edge device, our model achieves a processing speed of approximately from 14 to 39 FPS while maintaining a desirable mean Average Precision (mAP), which is capable of real-time detection on resource-constrained platforms.

STCA-SNN: self-attention-based temporal-channel joint attention for spiking neural networks.

Spiking Neural Networks (SNNs) have shown great promise in processing spatio-temporal information compared to Artificial Neural Networks (ANNs). However, there remains a performance gap between SNNs and ANNs, which impedes the practical application of SNNs. With intrinsic event-triggered property and temporal dynamics, SNNs have the potential to effectively extract spatio-temporal features from event streams. To leverage the temporal potential of SNNs, we propose a self-attention-based temporal-channel joint attention SNN (STCA-SNN) with end-to-end training, which infers attention weights along both temporal and channel dimensions concurrently. It models global temporal and channel information correlations with self-attention, enabling the network to learn 'what' and 'when' to attend simultaneously. Our experimental results show that STCA-SNNs achieve better performance on N-MNIST (99.67%), CIFAR10-DVS (81.6%), and N-Caltech 101 (80.88%) compared with the state-of-the-art SNNs. Meanwhile, our ablation study demonstrates that STCA-SNNs improve the accuracy of event stream classification tasks.

Efficient human activity recognition with spatio-temporal spiking neural networks.

In this study, we explore Human Activity Recognition (HAR), a task that aims to predict individuals' daily activities utilizing time series data obtained from wearable sensors for health-related applications. Although recent research has predominantly employed end-to-end Artificial Neural Networks (ANNs) for feature extraction and classification in HAR, these approaches impose a substantial computational load on wearable devices and exhibit limitations in temporal feature extraction due to their activation functions. To address these challenges, we propose the application of Spiking Neural Networks (SNNs), an architecture inspired by the characteristics of biological neurons, to HAR tasks. SNNs accumulate input activation as presynaptic potential charges and generate a binary spike upon surpassing a predetermined threshold. This unique property facilitates spatio-temporal feature extraction and confers the advantage of low-power computation attributable to binary spikes. We conduct rigorous experiments on three distinct HAR datasets using SNNs, demonstrating that our approach attains competitive or superior performance relative to ANNs, while concurrently reducing energy consumption by up to 94%.

Hybrid Spiking Fully Convolutional Neural Network for Semantic Segmentation

The spiking neural network (SNN) exhibits distinct advantages in terms of low power consumption due to its event-driven nature. However, it is limited to simple computer vision tasks because the direct training of SNNs is challenging. In this study, we propose a hybrid architecture called the spiking fully convolutional neural network (SFCNN) to expand the application of SNNs in the field of semantic segmentation. To train the SNN, we employ the surrogate gradient method along with backpropagation. The accuracy of mean intersection over union (mIoU) for the VOC2012 dataset is higher than that of existing spiking FCNs by almost 30%. The accuracy of mIoU can reach 39.6%. Moreover, the proposed hybrid SFCNN achieved excellent segmentation performance for other datasets such as COCO2017, DRIVE, and Cityscapes. Our hybrid SFCNN is a valuable and interesting contribution to extending the functionality of SNNs, especially for power-constrained applications.

Hybrid photonic deep convolutional residual spiking neural networks for text classification.

Spiking neural networks (SNNs) offer powerful computation capability due to its event-driven nature and temporal processing. However, it is still limited to shallow structure and simple tasks due to the training difficulty. In this work, we propose a deep convolutional residual spiking neural network (DCRSNN) for text classification tasks. In the DCRSNN, the feature extraction is achieved via a convolution SNN with residual connection, using the surrogate gradient direct training technique. Classification is performed by a fully-connected network. We also suggest a hybrid photonic DCRSNN, in which photonic SNNs are used for classification with a converted training method. The accuracy of hard and soft reset methods, as well as three different surrogate functions, were evaluated and compared across four different datasets. Results indicated a maximum accuracy of 76.36% for MR, 91.03% for AG News, 88.06% for IMDB and 93.99% for Yelp review polarity. Soft reset methods used in the deep convolutional SNN yielded slightly better accuracy than their hard reset counterparts. We also considered the effects of different pooling methods and observation time windows and found that the convergence accuracy achieved by convolutional SNNs was comparable to that of convolutional neural networks under the same conditions. Moreover, the hybrid photonic DCRSNN also shows comparable testing accuracy. This work provides new insights into extending the SNN applications in the field of text classification and natural language processing, which is interesting for the resources-restrained scenarios.

EdgeMap: An Optimized Mapping Toolchain for Spiking Neural Network in Edge Computing.

Spiking neural networks (SNNs) have attracted considerable attention as third-generation artificial neural networks, known for their powerful, intelligent features and energy-efficiency advantages. These characteristics render them ideally suited for edge computing scenarios. Nevertheless, the current mapping schemes for deploying SNNs onto neuromorphic hardware face limitations such as extended execution times, low throughput, and insufficient consideration of energy consumption and connectivity, which undermine their suitability for edge computing applications. To address these challenges, we introduce EdgeMap, an optimized mapping toolchain specifically designed for deploying SNNs onto edge devices without compromising performance. EdgeMap consists of two main stages. The first stage involves partitioning the SNN graph into small neuron clusters based on the streaming graph partition algorithm, with the sizes of neuron clusters limited by the physical neuron cores. In the subsequent mapping stage, we adopt a multi-objective optimization algorithm specifically geared towards mitigating energy costs and communication costs for efficient deployment. EdgeMap-evaluated across four typical SNN applications-substantially outperforms other state-of-the-art mapping schemes. The performance improvements include a reduction in average latency by up to 19.8%, energy consumption by 57%, and communication cost by 58%. Moreover, EdgeMap exhibits an impressive enhancement in execution time by a factor of 1225.44×, alongside a throughput increase of up to 4.02×. These results highlight EdgeMap's efficiency and effectiveness, emphasizing its utility for deploying SNN applications in edge computing scenarios.

Feasibility study on the application of a spiking neural network in myoelectric control systems.

In recent years, the effectiveness of a spiking neural network (SNN) for Electromyography (EMG) pattern recognition has been validated, but there is a lack of comprehensive consideration of the problems of heavy training burden, poor robustness, and high energy consumption in the application of actual myoelectric control systems. In order to explore the feasibility of the application of SNN in actual myoelectric control systems, this paper investigated an EMG pattern recognition scheme based on SNN. To alleviate the differences in EMG distribution caused by electrode shifts and individual differences, the adaptive threshold encoding was applied to gesture sample encoding. To improve the feature extraction ability of SNN, the leaky-integrate-and-fire (LIF) neuron that combines voltage-current effect was adopted as a spike neuron model. To balance recognition accuracy and power consumption, experiments were designed to determine encoding parameter and LIF neuron release threshold. By conducting the gesture recognition experiments considering different training test ratios, electrode shifts, and user independences on the nine-gesture high-density and low-density EMG datasets respectively, the advantages of the proposed SNN-based scheme have been verified. Compared with a Convolutional Neural Network (CNN), Long Short-Term Memory Network (LSTM) and Linear Discriminant Analysis (LDA), SNN can effectively reduce the number of repetitions in the training set, and its power consumption was reduced by 1-2 orders of magnitude. For the high-density and low-density EMG datasets, SNN improved the overall average accuracies by about (0.99 ~ 14.91%) under different training test ratios. For the high-density EMG dataset, the accuracy of SNN was improved by (0.94 ~ 13.76%) under electrode-shift condition and (3.81 ~ 18.95%) in user-independent case. The advantages of SNN in alleviating the user training burden, reducing power consumption, and improving robustness are of great significance for the implementation of user-friendly low-power myoelectric control systems.

Comparing Reservoir Artificial and Spiking Neural Networks in Machine Fault Detection Tasks

For the last two decades, artificial neural networks (ANNs) of the third generation, also known as spiking neural networks (SNN), have remained a subject of interest for researchers. A significant difficulty for the practical application of SNNs is their poor suitability for von Neumann computer architecture, so many researchers are currently focusing on the development of alternative hardware. Nevertheless, today several experimental libraries implementing SNNs for conventional computers are available. In this paper, using the RCNet library, we compare the performance of reservoir computing architectures based on artificial and spiking neural networks. We explicitly show that, despite the higher execution time, SNNs can demonstrate outstanding classification accuracy in the case of complicated datasets, such as data from industrial sensors used for the fault detection of bearings and gears. For one of the test problems, namely, ball bearing diagnosis using an accelerometer, the accuracy of the classification using reservoir SNN almost reached 100%, while the reservoir ANN was able to achieve recognition accuracy up to only 61%. The results of the study clearly demonstrate the superiority and benefits of SNN classificators.

A Compact Artificial Spiking Neuron Using a Sharp-Switching FET With Ultra-Low Energy Consumption Down to 0.45 fJ/Spike

In this article, we first utilize a zero subthreshold swing and zero impact ionization FET (Z2-FET) as an innovative artificial spiking neuron and demonstrate it with CMOS-compatible technology. Owing to the sharp-switching and hysteresis characteristics of Z2-FET, the artificial neuron successfully emulates the key biological neuronal behaviors based on a single device, including threshold-driven spiking and stimulus strength-modulated frequency response. Furthermore, the firing threshold of the neuron is conveniently tuned by the gate voltage, which is helpful in the application of spiking neural networks (SNN). A preliminary endurance up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${2}\times {10}^{{8}}$ </tex-math></inline-formula> cycles is obtained in the neuron. TCAD simulations further verify the scaling capability of the Z2-FET neuron and its energy consumption is estimated. The results suggest that <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text{Z}^{\vphantom {D^{f}}{2}}$ </tex-math></inline-formula> -FET has great potential for highly compact and energy-efficient artificial spiking neurons.

Synaptic Characteristic of Hafnia-Based Ferroelectric Tunnel Junction Device for Neuromorphic Computing Application.

Owing to the 4th Industrial Revolution, the amount of unstructured data, such as voice and video data, is rapidly increasing. Brain-inspired neuromorphic computing is a new computing method that can efficiently and parallelly process rapidly increasing data. Among artificial neural networks that mimic the structure of the brain, the spiking neural network (SNN) is a network that imitates the information-processing method of biological neural networks. Recently, memristors have attracted attention as synaptic devices for neuromorphic computing systems. Among them, the ferroelectric doped-HfO2-based ferroelectric tunnel junction (FTJ) is considered as a strong candidate for synaptic devices due to its advantages, such as complementary metal-oxide-semiconductor device/process compatibility, a simple two-terminal structure, and low power consumption. However, research on the spiking operations of FTJ devices for SNN applications is lacking. In this study, the implementation of long-term depression and potentiation as the spike timing-dependent plasticity (STDP) rule in the FTJ device was successful. Based on the measured data, a CrossSim simulator was used to simulate the classification of handwriting images. With a high accuracy of 95.79% for the Mixed National Institute of Standards and Technology (MNIST) dataset, the simulation results demonstrate that our device is capable of differentiating between handwritten images. This suggests that our FTJ device can be used as a synaptic device for implementing an SNN.

IC-SNN: Optimal ANN2SNN Conversion at Low Latency

The spiking neural network (SNN) has attracted the attention of many researchers because of its low energy consumption and strong bionics. However, when the network conversion method is used to solve the difficulty of network training caused by its discrete, too-long inference time, it may hinder the practical application of SNN. This paper proposes a novel model named the SNN with Initialized Membrane Potential and Coding Compensation (IC-SNN) to solve this problem. The model focuses on the effect of residual membrane potential and rate encoding on the target SNN. After analyzing the conversion error and the information loss caused by the encoding method under the low time step, we propose a new initial membrane potential setting method and coding compensation scheme. The model can enable the network to still achieve high accuracy under a low number of time steps by eliminating residual membrane potential and encoding errors in the SNN. Finally, experimental results based on public datasets CIFAR10 and CIFAR100 also demonstrate that the model can still achieve competitive classification accuracy in 32 time steps.

Beyond classification: directly training spiking neural networks for semantic segmentation

Spiking neural networks (SNNs) have recently emerged as the low-power alternative to artificial neural networks (ANNs) because of their sparse, asynchronous, and binary event-driven processing. Due to their energy efficiency, SNNs have a high possibility of being deployed for real-world, resource-constrained systems such as autonomous vehicles and drones. However, owing to their non-differentiable and complex neuronal dynamics, most previous SNN optimization methods have been limited to image recognition. In this paper, we explore the SNN applications beyond classification and present semantic segmentation networks configured with spiking neurons. Specifically, we first investigate two representative SNN optimization techniques for recognition tasks (i.e., ANN-SNN conversion and surrogate gradient learning) on semantic segmentation datasets. We observe that, when converted from ANNs, SNNs suffer from high latency and low performance due to the spatial variance of features. Therefore, we directly train networks with surrogate gradient learning, resulting in lower latency and higher performance than ANN-SNN conversion. Moreover, we redesign two fundamental ANN segmentation architectures (i.e., Fully Convolutional Networks and DeepLab) for the SNN domain. We conduct experiments on three semantic segmentation benchmarks including PASCAL VOC2012 dataset, DDD17 event-based dataset, and synthetic segmentation dataset combined CIFAR10 and MNIST datasets. In addition to showing the feasibility of SNNs for semantic segmentation, we show that SNNs can be more robust and energy-efficient compared to their ANN counterparts in this domain.

A Novel Window Function Enables Memristor Model With High Efficiency Spiking Neural Network Applications

A Novel Window Function Enables Memristor Model With High Efficiency Spiking Neural Network Applications

Optimized Potential Initialization for Low-Latency Spiking Neural Networks

Spiking Neural Networks (SNNs) have been attached great importance due to the distinctive properties of low power consumption, biological plausibility, and adversarial robustness. The most effective way to train deep SNNs is through ANN-to-SNN conversion, which have yielded the best performance in deep network structure and large-scale datasets. However, there is a trade-off between accuracy and latency. In order to achieve high precision as original ANNs, a long simulation time is needed to match the firing rate of a spiking neuron with the activation value of an analog neuron, which impedes the practical application of SNN. In this paper, we aim to achieve high-performance converted SNNs with extremely low latency (fewer than 32 time-steps). We start by theoretically analyzing ANN-to-SNN conversion and show that scaling the thresholds does play a similar role as weight normalization. Instead of introducing constraints that facilitate ANN-to-SNN conversion at the cost of model capacity, we applied a more direct way by optimizing the initial membrane potential to reduce the conversion loss in each layer. Besides, we demonstrate that optimal initialization of membrane potentials can implement expected error-free ANN-to-SNN conversion. We evaluate our algorithm on the CIFAR-10 dataset and CIFAR-100 dataset and achieve state-of-the-art accuracy, using fewer time-steps. For example, we reach top-1 accuracy of 93.38% on CIFAR-10 with 16 time-steps. Moreover, our method can be applied to other ANN-SNN conversion methodologies and remarkably promote performance when the time-steps is small.

Recognition of Electromagnetic Signals Based on the Spiking Convolutional Neural Network

Feature extraction and recognition of signals are the bases of a cognitive radio. Traditional manual extraction for signals’ features becomes difficult in the complex electromagnetic environment. Although convolutional neural networks (CNNs) can extract signal features automatically, they have low accuracy in recognizing electromagnetic signals at low signal-to-noise ratios (SNRs) due to the agility of signals. Considering the great potential of spiking neural networks (SNNs) in classification, a spiking convolution neural network (SCNN) for the recognition of electromagnetic signals is proposed in this paper. The SCNN effectively integrates the extraction ability of spatial features in CNNs and temporal features in SNNs. Since the SCNN is difficult to train, the strategy of surrogate gradient is proposed to train it. By taking the 2-dimensional time-frequency distribution of 6 signals as input, the SCNN can effectively identify different signals at low SNRs. The method proposed in this paper contributes to promote the research and application of SNNs in the recognition of electromagnetic signals.