Research on low-resolution pedestrian detection algorithms based on R-CNN with targeted pooling and proposal

Peng Shi, Jun Wu, Kai Wang, Yao Zhang, Jiapei Wang, Juneho Yi

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

1 Scopus citations

Abstract

We present an effective low-resolution pedestrian detection using targeted pooling and Region Proposal Network (RPN) in the Faster R-CNN. Our method firstly rearranges the anchor from the RPN exploiting an optimal hyper-parameter setting called »Elaborate Setup». Secondly, it refines the granularity in the pooling operation from the ROI pooling layer. The experimental results demonstrate that the proposed RPN together with fine-grained pooling, which we call LRPD-R-CNN is able to achieve high average precision and robust performance on the VOC 2007 dataset. This method has great potential in commercial values and wide application prospect in the field of computer vision, security and intelligent city.

Original languageEnglish
Title of host publication2018 8th International Conference on Image Processing Theory, Tools and Applications, IPTA 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538664278
DOIs
StatePublished - 10 Jan 2019
Event8th International Conference on Image Processing Theory, Tools and Applications, IPTA 2018 - Xi'an, China
Duration: 7 Nov 201810 Nov 2018

Publication series

Name2018 8th International Conference on Image Processing Theory, Tools and Applications, IPTA 2018 - Proceedings

Conference

Conference8th International Conference on Image Processing Theory, Tools and Applications, IPTA 2018
Country/TerritoryChina
CityXi'an
Period7/11/1810/11/18

Keywords

  • Anchors
  • Faster R-CNN
  • Fine-grained Pooling
  • Low-Resolution Pedestrian Detection

Fingerprint

Dive into the research topics of 'Research on low-resolution pedestrian detection algorithms based on R-CNN with targeted pooling and proposal'. Together they form a unique fingerprint.

Cite this