Event-based Video Reconstruction via Potential-assisted Spiking Neural Network

Abstract

Neuromorphic vision sensor is a new bio-inspired imaging paradigm that reports asynchronous, continuously perpixel brightness changes called 'events' with high temporal resolution and high dynamic range. So far, the eventbased image reconstruction methods are based on artificial neural networks (ANN) or hand-crafted spatiotemporal smoothing techniques. In this paper, we first implement the image reconstruction work via fully spiking neural network (SNN) architecture. As the bio-inspired neural networks, SNNs operating with asynchronous binary spikes distributed over time, can potentially lead to greater computational efficiency on event-driven hardware. We propose a novel Event-based Video reconstruction framework based on a fully Spiking Neural Network (EVSNN), which utilizes Leaky-Integrate-and-Fire (LIF) neuron and Membrane Potential (MP) neuron. We find that the spiking neurons have the potential to store useful temporal information (memory) to complete such time-dependent tasks. Furthermore, to better utilize the temporal information, we propose a hybrid potential-assisted framework (PA-EVSNN) using the membrane potential of spiking neuron. The proposed neuron is referred as Adaptive Membrane Potential (AMP) neuron, which adaptively updates the membrane potential according to the input spikes. The experimental results demonstrate that our models achieve comparable performance to ANN-based models on IJRR, MVSEC, and HQF datasets. The energy consumptions of EVSNN and PA-EVSNN are 19.36× and 7.75× more computationally efficient than their ANN architectures, respectively. Our code and pretrained model will be released later.

Huawei Technologies

“…With improved training techniques such as specialized normalization, SG, and loss function design [57,62,34,11], directly trained SNNs are achieving competitive accuracy on hard benchmark classification tasks like ImageNet, with only a few simulation steps required for convergence. These advances have encouraged their application to other event-based vision tasks beyond classification, such as optical flow estimation [18], video reconstruction [67], and object detection [8]. Nevertheless, these works are still limited by relatively simple handcrafted architectures and are suboptimal in terms of accuracy compared to state-of-the-art ANNs.…”