Training Spiking Neural Networks Research Articles

Spiking Transfer Learning from RGB Image to Neuromorphic Event Stream.

Recent advances in bio-inspired vision with event cameras and associated spiking neural networks (SNNs) have provided promising solutions for low-power consumption neuromorphic tasks. However, as the research of event cameras is still in its infancy, the amount of labeled event stream data is much less than that of the RGB database. The traditional method of converting static images into event streams by simulation to increase the sample size cannot simulate the characteristics of event cameras such as high temporal resolution. To take advantage of both the rich knowledge in labeled RGB images and the features of the event camera, we propose a transfer learning method from the RGB to the event domain in this paper. Specifically, we first introduce a transfer learning framework named R2ETL (RGB to Event Transfer Learning), including a novel encoding alignment module and a feature alignment module. Then, we introduce the temporal centered kernel alignment (TCKA) loss function to improve the efficiency of transfer learning. It aligns the distribution of temporal neuron states by adding a temporal learning constraint. Finally, we theoretically analyze the amount of data required by the deep neuromorphic model to prove the necessity of our method. Numerous experiments demonstrate that our proposed framework outperforms the state-of-the-art SNN and artificial neural network (ANN) models trained on event streams, including N-MNIST, CIFAR10-DVS and N-Caltech101. This indicates that the R2ETL framework is able to leverage the knowledge of labeled RGB images to help the training of SNN on event streams.

LDD: High-Precision Training of Deep Spiking Neural Network Transformers Guided by an Artificial Neural Network

The rise of large-scale Transformers has led to challenges regarding computational costs and energy consumption. In this context, spiking neural networks (SNNs) offer potential solutions due to their energy efficiency and processing speed. However, the inaccuracy of surrogate gradients and feature space quantization pose challenges for directly training deep SNN Transformers. To tackle these challenges, we propose a method (called LDD) to align ANN and SNN features across different abstraction levels in a Transformer network. LDD incorporates structured feature knowledge from ANNs to guide SNN training, ensuring the preservation of crucial information and addressing inaccuracies in surrogate gradients through designing layer-wise distillation losses. The proposed approach outperforms existing methods on the CIFAR10 (96.1%), CIFAR100 (82.3%), and ImageNet (80.9%) datasets, and enables training of the deepest SNN Transformer network using ImageNet.

LDD: High-Precision Training of Deep Spiking Neural Network Transformers Guided by an Artificial Neural Network

The rise of large-scale Transformers has led to challenges regarding computational costs and energy consumption. In this context, spiking neural networks (SNNs) offer potential solutions due to their energy efficiency and processing speed. However, the inaccuracy of surrogate gradients and feature space quantization pose challenges for directly training deep SNN Transformers. To tackle these challenges, we propose a method (called LDD) to align ANN and SNN features across different abstraction levels in a Transformer network. LDD incorporates structured feature knowledge from ANNs to guide SNN training, ensuring the preservation of crucial information and addressing inaccuracies in surrogate gradients through designing layer-wise distillation losses. The proposed approach outperforms existing methods on the CIFAR10 (96.1%), CIFAR100 (82.3%), and ImageNet (80.9%) datasets, and enables training of the deepest SNN Transformer network using ImageNet.

SNNtrainer3D: Training Spiking Neural Networks Using a User-Friendly Application with 3D Architecture Visualization Capabilities

Spiking Neural Networks have gained significant attention due to their potential for energy efficiency and biological plausibility. However, the reduced number of user-friendly tools for designing, training, and visualizing Spiking Neural Networks hinders widespread adoption. This paper presents the SNNtrainer3D v1.0.0, a novel software application that addresses these challenges. The application provides an intuitive interface for designing Spiking Neural Networks architectures, with features such as dynamic architecture editing, allowing users to add, remove, and edit hidden layers in real-time. A key innovation is the integration of Three.js for three-dimensional visualization of the network structure, enabling users to inspect connections and weights and facilitating a deeper understanding of the model’s behavior. The application supports training on the Modified National Institute of Standards and Technology dataset and allows the downloading of trained weights for further use. Moreover, it lays the groundwork for future integration with physical memristor technology, positioning it as a crucial tool for advancing neuromorphic computing research. The advantages of the development process, technology stack, and visualization are discussed. The SNNtrainer3D represents a significant step in making Spiking Neural Networks more accessible, understandable, and easier for Artificial Intelligence researchers and practitioners.

Fully neuromorphic vision and control for autonomous drone flight.

Biological sensing and processing is asynchronous and sparse, leading to low-latency and energy-efficient perception and action. In robotics, neuromorphic hardware for event-based vision and spiking neural networks promises to exhibit similar characteristics. However, robotic implementations have been limited to basic tasks with low-dimensional sensory inputs and motor actions because of the restricted network size in current embedded neuromorphic processors and the difficulties of training spiking neural networks. Here, we present a fully neuromorphic vision-to-control pipeline for controlling a flying drone. Specifically, we trained a spiking neural network that accepts raw event-based camera data and outputs low-level control actions for performing autonomous vision-based flight. The vision part of the network, consisting of five layers and 28,800 neurons, maps incoming raw events to ego-motion estimates and was trained with self-supervised learning on real event data. The control part consists of a single decoding layer and was learned with an evolutionary algorithm in a drone simulator. Robotic experiments show a successful sim-to-real transfer of the fully learned neuromorphic pipeline. The drone could accurately control its ego-motion, allowing for hovering, landing, and maneuvering sideways-even while yawing at the same time. The neuromorphic pipeline runs on board on Intel's Loihi neuromorphic processor with an execution frequency of 200 hertz, consuming 0.94 watt of idle power and a mere additional 7 to 12 milliwatts when running the network. These results illustrate the potential of neuromorphic sensing and processing for enabling insect-sized intelligent robots.

Intelligent event-based lip reading word classification with spiking neural networks using spatio-temporal attention features and triplet loss

Lip reading is a visual alternative to enhance the intelligence of traditional speech recognition, which can benefit from retina-like event cameras that focus on dynamic movements. Spiking Neural Networks (SNNs) are inherently well-suited to cooperate with event cameras. However, employing SNN for event-based lip reading presents significant challenges, particularly in effectively extracting spatio-temporal features from input events and distinguishing between phonetically similar elements. To address these challenges, this paper proposes a novel event-based lip-reading model that leverages the SNN framework, enriched by a designed Spatial-Temporal Attention Block (STAB) and a triplet loss. Specifically, STAB comprises both spatial and temporal attention branches, dynamically emphasizing those spatial and temporal characteristics most reflective of lip movements. STAB also includes a fusion mechanism to integrate these spatial and temporal insights, providing a comprehensive and focused representation of lip movements. In addition, we enhance the model by incorporating triplet loss into the SNN training process, further improving the model's ability to distinguish between visually similar words. Experimental results show the superior performance of our SNN and validate the effect of STAB and triplet loss. We also conduct the energy consumption analysis to affirm the model's energy efficiency.

A Synapse-Threshold Synergistic Learning Approach for Spiking Neural Networks

Spiking neural networks (SNNs) have demonstrated excellent capabilities in various intelligent scenarios. Most existing methods for training SNNs are based on the concept of synaptic plasticity; however, learning in the realistic brain also utilizes intrinsic non-synaptic mechanisms of neurons. The spike threshold of biological neurons is a critical intrinsic neuronal feature that exhibits rich dynamics on a millisecond timescale and has been proposed as an underlying mechanism that facilitates neural information processing. In this study, we develop a novel synergistic learning approach that involves simultaneously training synaptic weights and spike thresholds in SNNs. SNNs trained with synapse-threshold synergistic learning (STL-SNNs) achieve significantly superior performance on various static and neuromorphic datasets than SNNs trained with two degenerated single-learning models. During training, the synergistic learning approach optimizes neural thresholds, providing the network with stable signal transmission via appropriate firing rates. Further analysis indicates that STL-SNNs are robust to noisy data and exhibit low energy consumption for deep network structures. Additionally, the performance of STL-SNN can be further improved by introducing a generalized joint decision framework. Overall, our findings indicate that biologically plausible synergies between synaptic and intrinsic non-synaptic mechanisms may provide a promising approach for developing highly efficient SNN learning methods.

Enhancing Training of Spiking Neural Network with Stochastic Latency

Spiking neural networks (SNNs) have garnered significant attention for their low power consumption when deployed on neuromorphic hardware that operates in orders of magnitude lower power than general-purpose hardware. Direct training methods for SNNs come with an inherent latency for which the SNNs are optimized, and in general, the higher the latency, the better the predictive powers of the models, but at the same time, the higher the energy consumption during training and inference. Furthermore, an SNN model optimized for one particular latency does not necessarily perform well in lower latencies, which becomes relevant in scenarios where it is necessary to switch to a lower latency because of the depletion of onboard energy or other operational requirements. In this work, we propose Stochastic Latency Training (SLT), a direct training method for SNNs that optimizes the model for the given latency but simultaneously offers a minimum reduction of predictive accuracy when shifted to lower inference latencies. We provide heuristics for our approach with partial theoretical justification and experimental evidence showing the state-of-the-art performance of our models on datasets such as CIFAR-10, DVS-CIFAR-10, CIFAR-100, and DVS-Gesture. Our code is available at https://github.com/srinuvaasu/SLT

Memory-Efficient Reversible Spiking Neural Networks

Spiking neural networks (SNNs) are potential competitors to artificial neural networks (ANNs) due to their high energy-efficiency on neuromorphic hardware. However, SNNs are unfolded over simulation time steps during the training process. Thus, SNNs require much more memory than ANNs, which impedes the training of deeper SNN models. In this paper, we propose the reversible spiking neural network to reduce the memory cost of intermediate activations and membrane potentials during training. Firstly, we extend the reversible architecture along temporal dimension and propose the reversible spiking block, which can reconstruct the computational graph and recompute all intermediate variables in forward pass with a reverse process. On this basis, we adopt the state-of-the-art SNN models to the reversible variants, namely reversible spiking ResNet (RevSResNet) and reversible spiking transformer (RevSFormer). Through experiments on static and neuromorphic datasets, we demonstrate that the memory cost per image of our reversible SNNs does not increase with the network depth. On CIFAR10 and CIFAR100 datasets, our RevSResNet37 and RevSFormer-4-384 achieve comparable accuracies and consume 3.79x and 3.00x lower GPU memory per image than their counterparts with roughly identical model complexity and parameters. We believe that this work can unleash the memory constraints in SNN training and pave the way for training extremely large and deep SNNs.

An Efficient Knowledge Transfer Strategy for Spiking Neural Networks from Static to Event Domain

Spiking neural networks (SNNs) are rich in spatio-temporal dynamics and are suitable for processing event-based neuromorphic data. However, event-based datasets are usually less annotated than static datasets. This small data scale makes SNNs prone to overfitting and limits their performance. In order to improve the generalization ability of SNNs on event-based datasets, we use static images to assist SNN training on event data. In this paper, we first discuss the domain mismatch problem encountered when directly transferring networks trained on static datasets to event data. We argue that the inconsistency of feature distributions becomes a major factor hindering the effective transfer of knowledge from static images to event data. To address this problem, we propose solutions in terms of two aspects: feature distribution and training strategy. Firstly, we propose a knowledge transfer loss, which consists of domain alignment loss and spatio-temporal regularization. The domain alignment loss learns domain-invariant spatial features by reducing the marginal distribution distance between the static image and the event data. Spatio-temporal regularization provides dynamically learnable coefficients for domain alignment loss by using the output features of the event data at each time step as a regularization term. In addition, we propose a sliding training strategy, which gradually replaces static image inputs probabilistically with event data, resulting in a smoother and more stable training for the network. We validate our method on neuromorphic datasets, including N-Caltech101, CEP-DVS, and N-Omniglot. The experimental results show that our proposed method achieves better performance on all datasets compared to the current state-of-the-art methods. Code is available at https://github.com/Brain-Cog-Lab/Transfer-for-DVS.

A chaotic spiking backpropagation approach to brain learning.

A highlight of the chaotic spiking backpropagation (CSBP) method, which is a powerful tool for directly training spiking neural networks and helps to understand the learning mechanisms of human brain.

A Deep Convolutional Spiking Neural Network for embedded applications

Deep neural networks (DNNs) have received a great deal of interest in solving everyday tasks in recent years. However, their computational and energy costs limit their use on mobile and edge devices. The neuromorphic computing approach called spiking neural networks (SNNs) represents a potential solution for bridging the gap between performance and computational expense. Despite the potential benefits of energy efficiency, the current SNNs are being used with datasets such as MNIST, Fashion-MNIST, and CIFAR10, limiting their applications compared to DNNs. Therefore, the applicability of SNNs to real-world applications, such as scene classification and forecasting epileptic seizures, must be demonstrated yet. This paper develops a deep convolutional spiking neural network (DCSNN) for embedded applications. We explore a convolutional architecture, Visual Geometry Group (VGG16), to implement deeper SNNs. To train a spiking model, we convert the pre-trained VGG16 into corresponding spiking equivalents with nearly comparable performance to the original one. The trained weights of VGG16 were then transferred to the equivalent SNN architecture while performing a proper weight–threshold balancing. The model is evaluated in two case studies: land use and land cover classification, and epileptic seizure detection. Experimental results show a classification accuracy of 94.88%, and seizure detection specificity of 99.45% and a sensitivity of 95.06%. It is confirmed that conversion-based training SNNs are promising, and the benefits of DNNs, such as solving complex and real-world problems, become available to SNNs.

Brain-inspired chaotic spiking backpropagation.

Spiking neural networks (SNNs) have superior energy efficiency due to their spiking signal transmission, which mimics biological nervous systems, but they are difficult to train effectively. Although surrogate gradient-based methods offer a workable solution, trained SNNs frequently fall into local minima because they are still primarily based on gradient dynamics. Inspired by the chaotic dynamics in animal brain learning, we propose a chaotic spiking backpropagation (CSBP) method that introduces a loss function to generate brain-like chaotic dynamics and further takes advantage of the ergodic and pseudo-random nature to make SNN learning effective and robust. From a computational viewpoint, we found that CSBP significantly outperforms current state-of-the-art methods on both neuromorphic data sets (e.g. DVS-CIFAR10 and DVS-Gesture) and large-scale static data sets (e.g. CIFAR100 and ImageNet) in terms of accuracy and robustness. From a theoretical viewpoint, we show that the learning process of CSBP is initially chaotic, then subject to various bifurcations and eventually converges to gradient dynamics, consistently with the observation of animal brain activity. Our work provides a superior core tool for direct SNN training and offers new insights into understanding the learning process of a biological brain.

Chip-In-Loop SNN Proxy Learning: a new method for efficient training of spiking neural networks.

The primary approaches used to train spiking neural networks (SNNs) involve either training artificial neural networks (ANNs) first and then transforming them into SNNs, or directly training SNNs using surrogate gradient techniques. Nevertheless, both of these methods encounter a shared challenge: they rely on frame-based methodologies, where asynchronous events are gathered into synchronous frames for computation. This strays from the authentic asynchronous, event-driven nature of SNNs, resulting in notable performance degradation when deploying the trained models on SNN simulators or hardware chips for real-time asynchronous computation. To eliminate this performance degradation, we propose a hardware-based SNN proxy learning method that is called Chip-In-Loop SNN Proxy Learning (CIL-SPL). This approach effectively eliminates the performance degradation caused by the mismatch between synchronous and asynchronous computations. To demonstrate the effectiveness of our method, we trained models using public datasets such as N-MNIST and tested them on the SNN simulator or hardware chip, comparing our results to those classical training methods.

Rethinking Pretraining as a Bridge From ANNs to SNNs.

Spiking neural networks (SNNs) are known as typical kinds of brain-inspired models with their unique features of rich neuronal dynamics, diverse coding schemes, and low power consumption properties. How to obtain a high-accuracy model has always been the main challenge in the field of SNN. Currently, there are two mainstream methods, i.e., obtaining a converted SNN through converting a well-trained artificial NN (ANN) to its SNN counterpart or training an SNN directly. However, the inference time of a converted SNN is too long, while SNN training is generally very costly and inefficient. In this work, a new SNN training paradigm is proposed by combining the concepts of the two different training methods with the help of the pretrain technique and BP-based deep SNN training mechanism. We believe that the proposed paradigm is a more efficient pipeline for training SNNs. The pipeline includes pipe-S for static data transfer tasks and pipe-D for dynamic data transfer tasks. State-of-the-art (SOTA) results are obtained in a large-scale event-driven dataset ES-ImageNet. For training acceleration, we achieve the same (or higher) best accuracy as similar leaky-integrate-and-fire (LIF)-SNNs using 1/8 training time on ImageNet-1K and 1/2 training time on ES-ImageNet and also provide a time-accuracy benchmark for a new dataset ES-UCF101. These experimental results reveal the similarity of the functions of parameters between ANNs and SNNs and also demonstrate various potential applications of this SNN training pipeline.

Effective Surrogate Gradient Learning With High-Order Information Bottleneck for Spike-Based Machine Intelligence.

Brain-inspired computing technique presents a promising approach to prompt the rapid development of artificial general intelligence (AGI). As one of the most critical aspects, spiking neural networks (SNNs) have demonstrated superiority for AGI, such as low power consumption. Effective training of SNNs with high generalization ability, high robustness, and low power consumption simultaneously is a significantly challenging problem for the development and success of applications of spike-based machine intelligence. In this research, we present a novel and flexible learning framework termed high-order spike-based information bottleneck (HOSIB) leveraging the surrogate gradient technique. The presented HOSIB framework, including second-order and third-order formation, i.e., second-order information bottleneck (SOIB) and third-order information bottleneck (TOIB), comprehensively explores the common latent architecture and the spike-based intrinsic information and discards the superfluous information in the data, which improves the generalization capability and robustness of SNN models. Specifically, HOSIB relies on the information bottleneck (IB) principle to prompt the sparse spike-based information representation and flexibly balance its exploitation and loss. Extensive classification experiments are conducted to empirically show the promising generalization ability of HOSIB. Furthermore, we apply the SOIB and TOIB algorithms in deep spiking convolutional networks to demonstrate their improvement in robustness with various categories of noise. The experimental results prove the HOSIB framework, especially TOIB, can achieve better generalization ability, robustness and power efficiency in comparison with the current representative studies.

Backpropagation-Based Learning Techniques for Deep Spiking Neural Networks: A Survey.

With the adoption of smart systems, artificial neural networks (ANNs) have become ubiquitous. Conventional ANN implementations have high energy consumption, limiting their use in embedded and mobile applications. Spiking neural networks (SNNs) mimic the dynamics of biological neural networks by distributing information over time through binary spikes. Neuromorphic hardware has emerged to leverage the characteristics of SNNs, such as asynchronous processing and high activation sparsity. Therefore, SNNs have recently gained interest in the machine learning community as a brain-inspired alternative to ANNs for low-power applications. However, the discrete representation of the information makes the training of SNNs by backpropagation-based techniques challenging. In this survey, we review training strategies for deep SNNs targeting deep learning applications such as image processing. We start with methods based on the conversion from an ANN to an SNN and compare these with backpropagation-based techniques. We propose a new taxonomy of spiking backpropagation algorithms into three categories, namely, spatial, spatiotemporal, and single-spike approaches. In addition, we analyze different strategies to improve accuracy, latency, and sparsity, such as regularization methods, training hybridization, and tuning of the parameters specific to the SNN neuron model. We highlight the impact of input encoding, network architecture, and training strategy on the accuracy-latency tradeoff. Finally, in light of the remaining challenges for accurate and efficient SNN solutions, we emphasize the importance of joint hardware-software codevelopment.

Efficient Deep Spiking Multilayer Perceptrons With Multiplication-Free Inference.

Advancements in adapting deep convolution architectures for spiking neural networks (SNNs) have significantly enhanced image classification performance and reduced computational burdens. However, the inability of multiplication-free inference (MFI) to align with attention and transformer mechanisms, which are critical to superior performance on high-resolution vision tasks, imposes limitations on these gains. To address this, our research explores a new pathway, drawing inspiration from the progress made in multilayer perceptrons (MLPs). We propose an innovative spiking MLP architecture that uses batch normalization (BN) to retain MFI compatibility and introduce a spiking patch encoding (SPE) layer to enhance local feature extraction capabilities. As a result, we establish an efficient multistage spiking MLP network that blends effectively global receptive fields with local feature extraction for comprehensive spike-based computation. Without relying on pretraining or sophisticated SNN training techniques, our network secures a top-one accuracy of 66.39% on the ImageNet-1K dataset, surpassing the directly trained spiking ResNet-34 by 2.67%. Furthermore, we curtail computational costs, model parameters, and simulation steps. An expanded version of our network compares with the performance of the spiking VGG-16 network with a 71.64% top-one accuracy, all while operating with a model capacity 2.1 times smaller. Our findings highlight the potential of our deep SNN architecture in effectively integrating global and local learning abilities. Interestingly, the trained receptive field in our network mirrors the activity patterns of cortical cells.

Design of spiking neural network architecture based on dendritic computation principles

The paper presents the hardware architecture design of a spiking neural network (SNN) based on dendritic computation principles. The integration of active dendritic properties into the neuronal structure of SNN aims to minimize the number of functional blocks required for hardware implementation, including synaptic connections and neurons. The available memory on the neuromorphic architecture imposes limitations on implementation, hence the need to reduce the number of functional blocks. As a test task for the SNN based on dendritic computations, we selected the image classification of eight symbols, consisting of digits one through eight. These symbols are depicted as 3×7 pixel, 1-bit images. Active dendritic properties were analyzed using the “delay plasticity” [1] principle, which introduces the mechanism of adjusting input signal delays in spiking neuron inputs. We designed an SNN model with complementary delay inputs, referred to as the active dendrite SNN, as a principle implementation. Input spikes arriving at the primary inputs are duplicated to the delay inputs after a modifiable time delay. For convenience, each delay input was set at a single value. The input images were scanned sequentially. The neural network received three direct and three inverse inputs from the six main inputs that were coded with spikes corresponding to three pixels of a string. An “on” pixel was coded with a spike arriving at a direct input, while an “off” pixel was coded with a spike arriving at the corresponding inverse input. The line scanning time was 10 μs, input width was 1 μs, and delay time was 5 μs. The optimization of spiking neuron parameters was performed through a stochastic search algorithm based on simulated annealing. The parameters optimized for the Leaky-Integrate-and-Fire (LIF) neurons included the leakage time constant (22.8 μs), firing threshold (1150 arbitrary units), and refractory period (1 μs). The active dendrite SNN training employed the tempotron learning rule [2]. The training optimized the following parameters: the maximum change in synaptic weight on potentiation and depression (0.7 and –3 arbitrary units, respectively) and the synaptic weight’s upper bound (195 arbitrary units). Complementary delayed inputs facilitated the learning of the order in which input patterns arrived for SNN neurons during training. The paper compares an SNN architecture based on dendritic computations to our previously designed two-layer SNN with a hidden perceptron layer and an output layer consisting of LIF neurons [3]. Using the same LIF neuron design, input image coding, and LIF neuron layer structure as in the proposed architecture, our two-layer SNN with a hidden perceptron layer and output layer of LIF neurons successfully recognized 3×5 images of three symbols with only 10 neurons and 63 synapses. Alternatively, the active dendrite SNN was able to recognize 3×7 images of eight symbols with four neurons and 48 synaptic weights. In conclusion, incorporating active dendrite properties into the SNN architecture for image recognition resulted in optimized functional block usage, lowering the number of neurons and synapses used by 60 and 24%, respectively.

Learnable Leakage and Onset-Spiking Self-Attention in SNNs with Local Error Signals.

Spiking neural networks (SNNs) have garnered significant attention due to their computational patterns resembling biological neural networks. However, when it comes to deep SNNs, how to focus on critical information effectively and achieve a balanced feature transformation both temporally and spatially becomes a critical challenge. To address these challenges, our research is centered around two aspects: structure and strategy. Structurally, we optimize the leaky integrate-and-fire (LIF) neuron to enable the leakage coefficient to be learnable, thus making it better suited for contemporary applications. Furthermore, the self-attention mechanism is introduced at the initial time step to ensure improved focus and processing. Strategically, we propose a new normalization method anchored on the learnable leakage coefficient (LLC) and introduce a local loss signal strategy to enhance the SNN's training efficiency and adaptability. The effectiveness and performance of our proposed methods are validated on the MNIST, FashionMNIST, and CIFAR-10 datasets. Experimental results show that our model presents a superior, high-accuracy performance in just eight time steps. In summary, our research provides fresh insights into the structure and strategy of SNNs, paving the way for their efficient and robust application in practical scenarios.