Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection

Object detection becomes a challenge due to diversity of object scales. In general, modern object detectors use feature pyramid to learn multi-scale representation for better results. However, current versions of feature pyramid are insufficient to handle scale imbalance, as it is inefficient to int...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Lubna Aziz, Md Sah Bin Haji Salam, Usman Ullah Sheikh, Surat Khan, Huma Ayub, Sara Ayub
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Object detection feature pyramid convolutional neural network computer vision Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/a3cdd6daffe54b7a8d1e2056afb40d3a
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:a3cdd6daffe54b7a8d1e2056afb40d3a
record_format	dspace
spelling	oai:doaj.org-article:a3cdd6daffe54b7a8d1e2056afb40d3a2021-12-02T00:00:41ZMulti-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection2169-353610.1109/ACCESS.2021.3130129https://doaj.org/article/a3cdd6daffe54b7a8d1e2056afb40d3a2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9624958/https://doaj.org/toc/2169-3536Object detection becomes a challenge due to diversity of object scales. In general, modern object detectors use feature pyramid to learn multi-scale representation for better results. However, current versions of feature pyramid are insufficient to handle scale imbalance, as it is inefficient to integrate semantic information across different scales. Here, we reformulate feature pyramid construction as a feature reconfiguration process. We propose a detection network, Multi-level Refinement Feature pyramid Network, to combine high-level features (i.e., semantic information), middle-level feature and low-level feature (i.e., boundary information), in a highly-nonlinear yet efficient manner. A novel contextual features module is proposed, which consists of global attention and local reconfigurations. It efficiently gathers task-oriented contextual features across different scales and spatial locations (i.e., lightweight local reconfiguration and global attention). To evaluate significance of proposed model, we designed and trained end-to-end single stage detector called MRFDet by assimilating it into Single Shot Detector (SSD), and it achieved better detection performance compared to most recent single-stage objects detectors. MRFDet achieves an AP of 45.2 with MS-COCO and an improvement in <inline-formula> <tex-math notation="LaTeX">$mAP$ </tex-math></inline-formula> of 4.5% with VOC.Lubna AzizMd Sah Bin Haji SalamUsman Ullah SheikhSurat KhanHuma AyubSara AyubIEEEarticleObject detectionfeature pyramidconvolutional neural networkcomputer visionElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 156492-156506 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Object detection feature pyramid convolutional neural network computer vision Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Object detection feature pyramid convolutional neural network computer vision Electrical engineering. Electronics. Nuclear engineering TK1-9971 Lubna Aziz Md Sah Bin Haji Salam Usman Ullah Sheikh Surat Khan Huma Ayub Sara Ayub Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
description	Object detection becomes a challenge due to diversity of object scales. In general, modern object detectors use feature pyramid to learn multi-scale representation for better results. However, current versions of feature pyramid are insufficient to handle scale imbalance, as it is inefficient to integrate semantic information across different scales. Here, we reformulate feature pyramid construction as a feature reconfiguration process. We propose a detection network, Multi-level Refinement Feature pyramid Network, to combine high-level features (i.e., semantic information), middle-level feature and low-level feature (i.e., boundary information), in a highly-nonlinear yet efficient manner. A novel contextual features module is proposed, which consists of global attention and local reconfigurations. It efficiently gathers task-oriented contextual features across different scales and spatial locations (i.e., lightweight local reconfiguration and global attention). To evaluate significance of proposed model, we designed and trained end-to-end single stage detector called MRFDet by assimilating it into Single Shot Detector (SSD), and it achieved better detection performance compared to most recent single-stage objects detectors. MRFDet achieves an AP of 45.2 with MS-COCO and an improvement in <inline-formula> <tex-math notation="LaTeX">$mAP$ </tex-math></inline-formula> of 4.5% with VOC.
format	article
author	Lubna Aziz Md Sah Bin Haji Salam Usman Ullah Sheikh Surat Khan Huma Ayub Sara Ayub
author_facet	Lubna Aziz Md Sah Bin Haji Salam Usman Ullah Sheikh Surat Khan Huma Ayub Sara Ayub
author_sort	Lubna Aziz
title	Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
title_short	Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
title_full	Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
title_fullStr	Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
title_full_unstemmed	Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection
title_sort	multi-level refinement feature pyramid network for scale imbalance object detection
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/a3cdd6daffe54b7a8d1e2056afb40d3a
work_keys_str_mv	AT lubnaaziz multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection AT mdsahbinhajisalam multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection AT usmanullahsheikh multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection AT suratkhan multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection AT humaayub multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection AT saraayub multilevelrefinementfeaturepyramidnetworkforscaleimbalanceobjectdetection
_version_	1718404004489199616

Multi-Level Refinement Feature Pyramid Network for Scale Imbalance Object Detection

Ejemplares similares