Event-based object detection with lightweight spatial attention mechanism

Zichen Liang, Guang Chen, Zhijun Li, Peigen Liu, Alois Knoll

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

Event camera conveys dynamic visual information in the format of asynchronous digital events, resulting to the disability of detectors developed for RGB images. Previous methods of event-based object detection mainly rely on simple template matching and encoded maps with deep learning, which sacrifices the spatial sparsity of events and achieves a weak performance in the noisy environment. This paper proposes a miniature event-based spatial attention mechanism of the one-stage detector to reduce the noise of events and to enrich the multi-scale feature maps by merging the shallow features. Maintaining the sparse property of events to the maximum degree, this paper transplants the model from convolution neural network to sparse convolution network and trains it in two ways (by its own and with knowledge distillation). Results show that the lightweight spatial attention mechanism is compatible with one-stage detectors and convolution neural network outperforms sparse convolution network in the event-based object detection.

Original languageEnglish
Title of host publication2021 6th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages498-503
Number of pages6
ISBN (Electronic)9780738133645
DOIs
StatePublished - 3 Jul 2021
Event6th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2021 - Chongqing, China
Duration: 3 Jul 20215 Jul 2021

Publication series

Name2021 6th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2021

Conference

Conference6th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2021
Country/TerritoryChina
CityChongqing
Period3/07/215/07/21

Fingerprint

Dive into the research topics of 'Event-based object detection with lightweight spatial attention mechanism'. Together they form a unique fingerprint.

Cite this