ScieNet: Deep Learning with Spike-assisted Contextual Information Extraction

Abstract

Deep neural networks (DNNs) provide high image classification accuracy, but experience significant performance degradation when perturbation from various sources are present in the input. The lack of resilience to input perturbations makes DNN less reliable for systems interacting with physical world such as autonomous vehicles, robotics, to name a few, where imperfect input is the normal condition. We present a hybrid deep network architecture with spike-assisted contextual information extraction (ScieNet). ScieNet integrates unsupervised learning using spiking neural network (SNN) for unsupervised contextual informationextraction with a back-end DNN trained for classification. The integrated network demonstrates high resilience to input perturbations without relying on prior training on perturbed inputs. We demonstrate ScieNet with different back-end DNNs for image classification using CIFAR dataset considering stochastic (noise) and structured (rain) input perturbations. Experimental results demonstrate significant improvement in accuracy on noisy and rainy images without prior training, while maintaining state-of-the-art accuracy on clean images.

References in corpus (5)

• Densely Connected Convolutional Networks
• Deep Learning in Spiking Neural Networks
• Bio-Inspired Spiking Convolutional Neural Network using Layer-wise Sparse Coding and STDP Learning
• Encoding Spike Patterns in Multilayer Spiking Neural Networks
• Mixture of Pre-processing Experts Model for Noise Robust Deep Learning on Resource Constrained Platforms