REGION AVERAGE POOLING FOR CONTEXT-AWARE OBJECT DETECTION

Page view(s)
32
Checked on Mar 20, 2025
REGION AVERAGE POOLING FOR CONTEXT-AWARE OBJECT DETECTION
Title:
REGION AVERAGE POOLING FOR CONTEXT-AWARE OBJECT DETECTION
Journal Title:
2017 IEEE International Conference on Image Processing
DOI:
Publication URL:
Publication Date:
17 September 2017
Citation:
Abstract:
Object detection has been a key task in computer vision with deep convolutional neural networks being a significant per- former. We propose a method named Region Average Pooling that leverages object co-occurrence to improve object detec- tion performance. Given regions of interest in an image, our method augments object detection networks with pooled con- textual features from other regions of interest in the scene. We implement our scheme and evaluate it against the Pascal Visual Object Classes (VOC) 2007 [1, 2] and Microsoft Com- mon Objects in Context (MS COCO) [3] datasets. When used as part of the Faster R-CNN object detection framework with VGG-16 [4], we show an increase in mAP (0.5:0.95) from 24.2% to 25.5% over baseline Faster R-CNN and Global Av- erage Pooling when testing on MS COCO.
License type:
PublisherCopyrights
Funding Info:
Description:
© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
ISBN:

Files uploaded:

File Size Format Action
region-average-pooling-kingsley.pdf 941.41 KB PDF Open