KLI

Stair-Step Feature Pyramid Networks for Object Detection

Metadata Downloads
Abstract
Feature Pyramid Networks have solved scale variation problems in object detection by developing multi-level features with different scales from backbone networks. Although this network achieved promising performance without affecting model complexity, they still suffer feature-level imbalance between multi-level features, i.e., low-level features and high-level features in each stage of the backbone. Moreover, the detection head predicts classification scores and offset regression independently on each feature of multi-level features, which leads to inconsistency among the detection branch. Hence, this paper releases this problem by introducing simple but effective Stair-step Feature Pyramid Networks (SFPN) to harmonize information between multi-level features. Further, the Offset Adaption Module (OA Module) is proposed to improve feature representation by adapting the feature of the classification branch with regressed offsets of the regression branch. On the MS-COCO dataset, the proposed method increases by 1.2% Average Precision when comparing with baseline FCOS without bells and whistles.
Author(s)
보 수언 투이쩐 띠엔 닷웬 주이 린조강현
Issued Date
2021
Type
Article
Keyword
Object detectionOffset AdaptionStair-step FPN
DOI
10.1007/978-3-030-81638-4_13
URI
https://oak.ulsan.ac.kr/handle/2021.oak/9156
https://ulsan-primo.hosted.exlibrisgroup.com/primo-explore/fulldisplay?docid=TN_cdi_springer_books_10_1007_978_3_030_81638_4_13&context=PC&vid=ULSAN&lang=ko_KR&search_scope=default_scope&adaptor=primo_central_multiple_fe&tab=default_tab&query=any,contains,Stair-Step%20Feature%20Pyramid%20Networks%20for%20Object%20Detection&offset=0&pcAvailability=true
Publisher
Communications in Computer and Information Science
Location
스위스
Language
영어
ISSN
1865-0929
Citation Volume
1405
Citation Number
1
Citation Start Page
168
Citation End Page
175
Appears in Collections:
Engineering > IT Convergence
Authorize & License
  • Authorize공개
Files in This Item:
  • There are no files associated with this item.

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.