Close
Web of Science
Page
1
(
Records
1
--
1)
[
1
]
Print
Record
1
of
1
Title
:
ASIF
-
Net
:
Attention Steered Interweave Fusion Network for RGB
-
D Salient Object Detection
Author
(
s
):
Li
,
CY
(
Li
,
Chongyi
);
Cong
,
RM
(
Cong
,
Runmin
);
Kwong
,
S
(
Kwong
,
Sam
);
Hou
,
JH
(
Hou
,
Junhui
);
Fu
,
HZ
(
Fu
,
Huazhu
);
Zhu
,
GP
(
Zhu
,
Guopu
);
Zhang
,
DW
(
Zhang
,
Dingwen
);
Huang
,
QM
(
Huang
,
Qingming
)
Source
:
IEEE TRANSACTIONS ON CYBERNETICS
Volume
:
51
Issue
:
1
Pages
:
88-100
DOI
:
10.1109/
TCYB
.2020.2969255
Published
:
JAN
2021
Times Cited in Web of Science Core Collection
:
16
Total Times Cited
:
17
Usage Count
(
Last
180
days
):
15
Usage Count
(
Since
2013):
15
Cited Reference Count
:
75
Abstract
:
Salient object detection from RGB
-
D images is an important yet challenging vision task
,
which aims at detecting the most distinctive objects in a
scene by combining color information and depth constraints
.
Unlike prior fusion manners
,
we propose an attention steered interweave fusion network
(
ASIF
-
Net
)
to detect salient objects
,
which progressively integrates cross
-
modal and cross
-
level complementarity from the RGB image and corresponding depth
map via steering of an attention mechanism
.
Specifically
,
the complementary features from RGB
-
D images are jointly extracted and hierarchically fused in a
dense and interweaved manner
.
Such a manner breaks down the barriers of inconsistency existing in the cross
-
modal data and also su
ff
iciently captures the
complementarity
.
Meanwhile
,
an attention mechanism is introduced to locate the potential salient regions in an attention
-
weighted fashion
,
which advances
in highlighting the salient objects and suppressing the cluttered background regions
.
Instead of focusing only on pixelwise saliency
,
we also ensure that the
detected salient objects have the objectness characteristics
(
e
.
g
.,
complete structure and sharp boundary
)
by incorporating the adversarial learning that
provides a global semantic constraint for RGB
-
D salient object detection
.
Quantitative and qualitative experiments demonstrate that the proposed method
performs favorably against
17
state
-
of
-
the
-
art saliency detectors on four publicly available RGB
-
D salient object detection datasets
.
The code and results of
our method are available at https
://
github
.
com
/
Li
-
Chongyi
/
ASIF
-
Net
.
Accession Number
:
WOS
:000602709000008
PubMed ID
:
32078571
Language
:
English
Document Type
:
Article
Author Keywords
:
Feature extraction
;
Saliency detection
;
Object detection
;
Task analysis
;
Fuses
;
Random access memory
;
Semantics
;
Adversarial learning
;
depth cue
;
interweave fusion
;
residual attention
;
RGB
-
D images
;
saliency detection
KeyWords Plus
:
OPTIMIZATION
Addresses
:
[
Li
,
Chongyi
;
Kwong
,
Sam
;
Hou
,
Junhui
]
City Univ Hong Kong
,
Dept Comp Sci
,
Hong Kong
,
Peoples R China
.
[
Cong
,
Runmin
]
Beijing Jiaotong Univ
,
Inst Informat Sci
,
Beijing
100044,
Peoples R China
.
[
Cong
,
Runmin
]
Beijing Key Lab Adv Informat Sci
&
Network Techno
,
Beijing
100044,
Peoples R China
.
[
Kwong
,
Sam
;
Hou
,
Junhui
]
City Univ Hong Kong
,
Shenzhen Res Inst
,
Shenzhen
51800,
Peoples R China
.
[
Fu
,
Huazhu
]
Incept Inst Artificial Intelligence
,
Abu Dhabi
,
U Arab Emirates
.
[
Zhu
,
Guopu
]
Chinese Acad Sci
,
Shenzhen Inst Adv Technol
,
Shenzhen
518055,
Peoples R China
.
[
Zhang
,
Dingwen
]
Xidian Univ
,
Sch Mechanoelect Engn
,
Xian
710071,
Peoples R China
.
[
Huang
,
Qingming
]
Univ Chinese Acad Sci
,
Sch Comp Sci
&
Technol
,
Beijing
100190,
Peoples R China
.
Corresponding Address
:
Cong
,
RM
(
corresponding author
),
Beijing Jiaotong Univ
,
Inst Informat Sci
,
Beijing
100044,
Peoples R China
.
E
-
mail Addresses
:
lichongyi
25@
gmail
.
com
;
rmcong
@
bjtu
.
edu
.
cn
;
cssamk
@
cityu
.
edu
.
hk
;
jh
.
hou
@
cityu
.
edu
.
hk
;
hzfu
@
ieee
.
org
;
gp
.
zhu
@
siat
.
ac
.
cn
;
zhangdingwen
2006
yyy
@
gmail
.
com
;
qmhuang
@
ucas
.
ac
.
cn
Author Identifiers
:
Author Web of Science ResearcherID ORCID Number
Li
,
Chengrui
S
-1262-2019
Hou
,
Junhui
0000-0003-3431-2021
Zhang
,
Dingwen
S
-9447-2017
0000-0001-8369-8886
Fu
,
Huazhu
0000-0002-9702-5524
,
sam
C
-9319-2012
0000-0001-7484-7261
Publisher
:
IEEE
-
INST ELECTRICAL ELECTRONICS ENGINEERS INC
Publisher Address
:
445
HOES LANE
,
PISCATAWAY
,
NJ
08855-4141
USA
Web of Science Categories
:
Automation
&
Control Systems
;
Computer Science
,
Artificial Intelligence
;
Computer Science
,
Cybernetics
Research Areas
:
Automation
&
Control Systems
;
Computer Science
IDS Number
:
PK
8
TB
ISSN
:
2168-2267
eISSN
:
2168-2275
29-
char Source Abbrev
.:
IEEE T CYBERNETICS
ISO Source Abbrev
.:
IEEE T
.
Cybern
.
Source Item Page Count
:
13
Funding
:
Funding Agency Grant Number