Please use this identifier to cite or link to this item:
Title: Cross-image region mining with region prototypical network for weakly supervised segmentation
Authors: Liu, Weide
Kong, Xiangfei
Hung, Tzu-Yi
Lin, Guosheng
Keywords: Engineering::Computer science and engineering
Issue Date: 2021
Source: Liu, W., Kong, X., Hung, T. & Lin, G. (2021). Cross-image region mining with region prototypical network for weakly supervised segmentation. IEEE Transactions On Multimedia, 3139459-.
Project: SMA-RP10
RG28/18 (S)
RG22/19 (S)
Journal: IEEE Transactions on Multimedia
Abstract: Weakly supervised image segmentation trained with image-level labels usually suffers from inaccurate coverage of object areas during the generation of the pseudo groundtruth. This is because the object activation maps are trained with the classification objective and lack the ability to generalize. To improve the generality of the objective activation maps, we propose a region prototypical network (RPNet) to explore the cross-image object diversity of the training set. Similar object parts across images are identified via region feature comparison. Object confidence is propagated between regions to discover new object areas while background regions are suppressed. Experiments show that the proposed method generates more complete and accurate pseudo object masks, while achieving state-of-the-art performance on PASCAL VOC 2012 and MS COCO. In addition, we investigate the robustness of the proposed method on reduced training sets.
ISSN: 1520-9210
DOI: 10.1109/TMM.2021.3139459
Rights: © 2021 IEEE. All rights reserved.
Fulltext Permission: none
Fulltext Availability: No Fulltext
Appears in Collections:SCSE Journal Articles

Citations 50

Updated on Mar 18, 2023

Page view(s)

Updated on Mar 23, 2023

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.