EvGNN: An Event-driven Graph Neural Network Accelerator for Edge Vision

arXiv:2404.19489v1 Announce Type: new
Abstract: Edge vision systems combining sensing and embedded processing promise low-latency, decentralized, and energy-efficient solutions that forgo reliance on the cloud. As opposed to conventional frame-based vision sensors, event-based cameras deliver a microsecond-scale temporal resolution with sparse information encoding, thereby outlining new opportunities for edge vision systems. However, mainstream algorithms for frame-based vision, which mostly rely on convolutional neural networks (CNNs), can hardly exploit the advantages of event-based vision as they are typically optimized for dense …

abstract accelerator arxiv cameras cloud cs.ar cs.cv cs.et cs.ne decentralized edge embedded encoding energy event graph graph neural network information latency low network neural network opportunities processing reliance resolution scale sensing sensors solutions systems temporal type vision