English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Event cameras attract researchers' attention due to their low powerconsumption, high dynamic range, and extremely high temporal resolution.Learning models on event-based object classification have recently achievedmassive success by accumulating sparse events into dense frames to applytraditional 2D learning methods. Yet, these approaches necessitate heavy-weightmodels and are with high computational complexity due to the redundantinformation introduced by the sparse-to-dense conversion, limiting thepotential of event cameras on real-life applications. This study aims toaddress the core problem of balancing accuracy and model complexity forevent-based classification models. To this end, we introduce a novel graphrepresentation for event data to exploit their sparsity better and customize alightweight voxel graph convolutional neural network (\textit{EV-VGCNN}) forevent-based classification. Specifically, (1) using voxel-wise vertices ratherthan previous point-wise inputs to explicitly exploit regional 2D semantics ofevent streams while keeping the sparsity;(2) proposing a multi-scale featurerelational layer (\textit{MFRL}) to extract spatial and motion cues from eachvertex discriminatively concerning its distances to neighbors. Comprehensiveexperiments show that our model can advance state-of-the-art classificationaccuracy with extremely low model complexity (merely 0.84M parameters).

A Voxel Graph CNN for Object Classification with Event Cameras