论文标题
具有时空压缩和突触卷积块的超低潜伏期尖峰神经网络
Ultra-low Latency Spiking Neural Networks with Spatio-Temporal Compression and Synaptic Convolutional Block
论文作者
论文摘要
作为脑灵感模型之一，尖峰神经网络（SNN）具有时空信息处理能力，低功率特征和高生物学上的合理性。有效的时空特征使其适用于事件流分类。但是，神经形态数据集，例如N-MNIST，CIFAR10-DVS，DVS128势头，需要将单个事件汇总为具有新的较高时间分辨率的新的事件分类，这会导致高训练和推断潜伏期。在这项工作中，我们提出了一种时空压缩方法，将单个事件汇总为突触电流的几个时间步骤，以减少训练和推理潜伏期。为了保持SNN在高压比下的准确性，我们还提出了一个突触卷积块，以平衡相邻时间步长之间的急剧变化。引入了具有可学习的膜时间常数的多阈值泄漏的集成和火力（LIF），以提高其信息处理能力。我们评估了关于神经形态N-MNIST，CIFAR10-DVS，DVS128手势数据集的事件流分类任务的建议方法。实验结果表明，我们所提出的方法使用更少的时间步骤优于几乎所有数据集的最先进的精度。
Spiking neural networks (SNNs), as one of the brain-inspired models, has spatio-temporal information processing capability, low power feature, and high biological plausibility. The effective spatio-temporal feature makes it suitable for event streams classification. However, neuromorphic datasets, such as N-MNIST, CIFAR10-DVS, DVS128-gesture, need to aggregate individual events into frames with a new higher temporal resolution for event stream classification, which causes high training and inference latency. In this work, we proposed a spatio-temporal compression method to aggregate individual events into a few time steps of synaptic current to reduce the training and inference latency. To keep the accuracy of SNNs under high compression ratios, we also proposed a synaptic convolutional block to balance the dramatic change between adjacent time steps. And multi-threshold Leaky Integrate-and-Fire (LIF) with learnable membrane time constant is introduced to increase its information processing capability. We evaluate the proposed method for event streams classification tasks on neuromorphic N-MNIST, CIFAR10-DVS, DVS128 gesture datasets. The experiment results show that our proposed method outperforms the state-of-the-art accuracy on nearly all datasets, using fewer time steps.
