Efficient Weakly-Supervised Object Detection With Pseudo Annotations
Weakly-supervised object detection (WSOD) has attracted lots of attention in recent years. However, there is still a big gap between WSOD and generic object detection. The main barriers to the efficiency of WSOD are the ineffective data augmentations and inaccurate bounding box predictions. Given on...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9494369/ |
id |
doaj-1375b99cee5a4957b280749c8d615890 |
---|---|
record_format |
Article |
spelling |
doaj-1375b99cee5a4957b280749c8d6158902021-07-29T23:00:37ZengIEEEIEEE Access2169-35362021-01-01910435610436610.1109/ACCESS.2021.30994979494369Efficient Weakly-Supervised Object Detection With Pseudo AnnotationsQingsheng Yuan0Gang Sun1Jianming Liang2https://orcid.org/0000-0002-9986-4436Biao Leng3https://orcid.org/0000-0003-3588-5622National Computer Network Emergency Response Technical Team/Coordination Center of China (CNCERT/CC), Beijing, ChinaThe School of Basic Education, Guangzhou Sports University, Guangzhou, ChinaSchool of Computer Science and Engineering, Beihang University, Beijing, ChinaSchool of Computer Science and Engineering, Beihang University, Beijing, ChinaWeakly-supervised object detection (WSOD) has attracted lots of attention in recent years. However, there is still a big gap between WSOD and generic object detection. The main barriers to the efficiency of WSOD are the ineffective data augmentations and inaccurate bounding box predictions. Given only image-level annotations, it is hard for WSOD to effectively utilize variant data augmentations and accurately regress the bounding boxes. Although a fully-supervised object detector can be trained using annotations generated from the weakly-supervised object detector, the performance is still severely limited due to the low quality of mined pseudo annotations. This paper proposes an efficient WSOD method with pseudo annotations (EWPA) to make better use of imperfect annotations. With the assistance of pseudo annotations, EWPA can effectively regress more accurate bounding boxes while the traditional WSOD can only locate the salient parts of an object. Furthermore, pseudo annotations can help design more complex data augmentations, driving the network to learn more discriminative feature representations. Extensive experiments are conducted on PASCAL VOC 2007 and 2012 datasets and validate the effectiveness of EWPA.https://ieeexplore.ieee.org/document/9494369/Object detectionweakly-supervised learningdata augmentationmixed-supervision |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Qingsheng Yuan Gang Sun Jianming Liang Biao Leng |
spellingShingle |
Qingsheng Yuan Gang Sun Jianming Liang Biao Leng Efficient Weakly-Supervised Object Detection With Pseudo Annotations IEEE Access Object detection weakly-supervised learning data augmentation mixed-supervision |
author_facet |
Qingsheng Yuan Gang Sun Jianming Liang Biao Leng |
author_sort |
Qingsheng Yuan |
title |
Efficient Weakly-Supervised Object Detection With Pseudo Annotations |
title_short |
Efficient Weakly-Supervised Object Detection With Pseudo Annotations |
title_full |
Efficient Weakly-Supervised Object Detection With Pseudo Annotations |
title_fullStr |
Efficient Weakly-Supervised Object Detection With Pseudo Annotations |
title_full_unstemmed |
Efficient Weakly-Supervised Object Detection With Pseudo Annotations |
title_sort |
efficient weakly-supervised object detection with pseudo annotations |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2021-01-01 |
description |
Weakly-supervised object detection (WSOD) has attracted lots of attention in recent years. However, there is still a big gap between WSOD and generic object detection. The main barriers to the efficiency of WSOD are the ineffective data augmentations and inaccurate bounding box predictions. Given only image-level annotations, it is hard for WSOD to effectively utilize variant data augmentations and accurately regress the bounding boxes. Although a fully-supervised object detector can be trained using annotations generated from the weakly-supervised object detector, the performance is still severely limited due to the low quality of mined pseudo annotations. This paper proposes an efficient WSOD method with pseudo annotations (EWPA) to make better use of imperfect annotations. With the assistance of pseudo annotations, EWPA can effectively regress more accurate bounding boxes while the traditional WSOD can only locate the salient parts of an object. Furthermore, pseudo annotations can help design more complex data augmentations, driving the network to learn more discriminative feature representations. Extensive experiments are conducted on PASCAL VOC 2007 and 2012 datasets and validate the effectiveness of EWPA. |
topic |
Object detection weakly-supervised learning data augmentation mixed-supervision |
url |
https://ieeexplore.ieee.org/document/9494369/ |
work_keys_str_mv |
AT qingshengyuan efficientweaklysupervisedobjectdetectionwithpseudoannotations AT gangsun efficientweaklysupervisedobjectdetectionwithpseudoannotations AT jianmingliang efficientweaklysupervisedobjectdetectionwithpseudoannotations AT biaoleng efficientweaklysupervisedobjectdetectionwithpseudoannotations |
_version_ |
1721247944003289088 |