Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism

Feng Qiu; Shuying Zhao; Yunzhou Zhang; Rong Ma; Yan Liu; Zhenyu Wang; Sonya Coleman

doi:10.1109/ICME52920.2022.9859674

Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism

Feng Qiu, Shuying Zhao, Yunzhou Zhang, Rong Ma, Yan Liu, Zhenyu Wang, Sonya Coleman

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

Deep learning based salient object detection methods have recently received significant attention. However, current methods still suffer from shortcomings such as informative background information being ignored which a significant problem for image saliency understanding. Additionally, it is also a challenge to suppress the noisy features in the network. By analyzing the difference between high-level and low-level features from ResNet-50, we utilize a Bilateral Feature Fusion (BFF) module to deal with the problem caused by ignoring informative background information. Benefitting from the BFF module, our proposed network can capture more meaningful foreground and background cues, which helps to get a more accurate saliency map. Moreover, we adopt a Score Sorting Attention (SSA) module which suppresses noisy and irrelevant features. Experimental results on five benchmark datasets demonstrate that our proposed method performs better than other state-of-the-art methods. The ablation studies also prove our contributions.

Original language	English
Title of host publication	Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME)
Publisher	IEEE Computer Society
ISBN (Electronic)	9781665485630
ISBN (Print)	978-1-6654-8564-7
DOIs	https://doi.org/10.1109/ICME52920.2022.9859674
Publication status	Published online - 26 Aug 2022
Event	2022 IEEE International Conference on Multimedia and Expo, ICME 2022 - Taipei, Taiwan, Republic of China Duration: 18 Jul 2022 → 22 Jul 2022

Publication series

Name	Proceedings - IEEE International Conference on Multimedia and Expo
Volume	2022-July
ISSN (Print)	1945-7871
ISSN (Electronic)	1945-788X

Conference

Conference	2022 IEEE International Conference on Multimedia and Expo, ICME 2022
Country/Territory	Taiwan, Republic of China
City	Taipei
Period	18/07/22 → 22/07/22

Bibliographical note

Funding Information:
∗Corresponding Author. This work was supported by Open Research Projects of Zhejiang Lab(No.2019KD0AD01/006), Major Science and Technology Projects of Liaoning Province(No.2021JH1/10400049), Fundation of Key Laboratory of Equipment Reliability(No.WD2C20205500306), Fundation of Key Laboratory of Aerospace System Simulation(No.6142002200301), National Natural Science Foundation of China (No. 61906176) and Fundamental Research Funds for the Central Universities (N2004022).

Publisher Copyright:
© 2022 IEEE.

Keywords

bilateral feature fusion
salient object detection
score sorting attention

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ICME52920.2022.9859674

Cite this

Qiu, F., Zhao, S., Zhang, Y., Ma, R., Liu, Y., Wang, Z., & Coleman, S. (2022). Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism. In Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME) Article 22012740 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2022-July). IEEE Computer Society. Advance online publication. https://doi.org/10.1109/ICME52920.2022.9859674

@inproceedings{ad3b3ea40d224f5180c00fcde909291c,

title = "Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism",

abstract = "Deep learning based salient object detection methods have recently received significant attention. However, current methods still suffer from shortcomings such as informative background information being ignored which a significant problem for image saliency understanding. Additionally, it is also a challenge to suppress the noisy features in the network. By analyzing the difference between high-level and low-level features from ResNet-50, we utilize a Bilateral Feature Fusion (BFF) module to deal with the problem caused by ignoring informative background information. Benefitting from the BFF module, our proposed network can capture more meaningful foreground and background cues, which helps to get a more accurate saliency map. Moreover, we adopt a Score Sorting Attention (SSA) module which suppresses noisy and irrelevant features. Experimental results on five benchmark datasets demonstrate that our proposed method performs better than other state-of-the-art methods. The ablation studies also prove our contributions.",

keywords = "bilateral feature fusion, salient object detection, score sorting attention",

author = "Feng Qiu and Shuying Zhao and Yunzhou Zhang and Rong Ma and Yan Liu and Zhenyu Wang and Sonya Coleman",

note = "Funding Information: ∗Corresponding Author. This work was supported by Open Research Projects of Zhejiang Lab(No.2019KD0AD01/006), Major Science and Technology Projects of Liaoning Province(No.2021JH1/10400049), Fundation of Key Laboratory of Equipment Reliability(No.WD2C20205500306), Fundation of Key Laboratory of Aerospace System Simulation(No.6142002200301), National Natural Science Foundation of China (No. 61906176) and Fundamental Research Funds for the Central Universities (N2004022). Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE International Conference on Multimedia and Expo, ICME 2022 ; Conference date: 18-07-2022 Through 22-07-2022",

year = "2022",

month = aug,

day = "26",

doi = "10.1109/ICME52920.2022.9859674",

language = "English",

isbn = "978-1-6654-8564-7",

series = "Proceedings - IEEE International Conference on Multimedia and Expo",

publisher = "IEEE Computer Society",

booktitle = "Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME)",

address = "United States",

}

Qiu, F, Zhao, S, Zhang, Y, Ma, R, Liu, Y, Wang, Z & Coleman, S 2022, Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism. in Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME)., 22012740, Proceedings - IEEE International Conference on Multimedia and Expo, vol. 2022-July, IEEE Computer Society, 2022 IEEE International Conference on Multimedia and Expo, ICME 2022, Taipei, Taiwan, Republic of China, 18/07/22. https://doi.org/10.1109/ICME52920.2022.9859674

Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism. / Qiu, Feng; Zhao, Shuying; Zhang, Yunzhou et al.
Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME). IEEE Computer Society, 2022. 22012740 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2022-July).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism

AU - Qiu, Feng

AU - Zhao, Shuying

AU - Zhang, Yunzhou

AU - Ma, Rong

AU - Liu, Yan

AU - Wang, Zhenyu

AU - Coleman, Sonya

N1 - Funding Information: ∗Corresponding Author. This work was supported by Open Research Projects of Zhejiang Lab(No.2019KD0AD01/006), Major Science and Technology Projects of Liaoning Province(No.2021JH1/10400049), Fundation of Key Laboratory of Equipment Reliability(No.WD2C20205500306), Fundation of Key Laboratory of Aerospace System Simulation(No.6142002200301), National Natural Science Foundation of China (No. 61906176) and Fundamental Research Funds for the Central Universities (N2004022). Publisher Copyright: © 2022 IEEE.

PY - 2022/8/26

Y1 - 2022/8/26

N2 - Deep learning based salient object detection methods have recently received significant attention. However, current methods still suffer from shortcomings such as informative background information being ignored which a significant problem for image saliency understanding. Additionally, it is also a challenge to suppress the noisy features in the network. By analyzing the difference between high-level and low-level features from ResNet-50, we utilize a Bilateral Feature Fusion (BFF) module to deal with the problem caused by ignoring informative background information. Benefitting from the BFF module, our proposed network can capture more meaningful foreground and background cues, which helps to get a more accurate saliency map. Moreover, we adopt a Score Sorting Attention (SSA) module which suppresses noisy and irrelevant features. Experimental results on five benchmark datasets demonstrate that our proposed method performs better than other state-of-the-art methods. The ablation studies also prove our contributions.

AB - Deep learning based salient object detection methods have recently received significant attention. However, current methods still suffer from shortcomings such as informative background information being ignored which a significant problem for image saliency understanding. Additionally, it is also a challenge to suppress the noisy features in the network. By analyzing the difference between high-level and low-level features from ResNet-50, we utilize a Bilateral Feature Fusion (BFF) module to deal with the problem caused by ignoring informative background information. Benefitting from the BFF module, our proposed network can capture more meaningful foreground and background cues, which helps to get a more accurate saliency map. Moreover, we adopt a Score Sorting Attention (SSA) module which suppresses noisy and irrelevant features. Experimental results on five benchmark datasets demonstrate that our proposed method performs better than other state-of-the-art methods. The ablation studies also prove our contributions.

KW - bilateral feature fusion

KW - salient object detection

KW - score sorting attention

UR - http://www.scopus.com/inward/record.url?scp=85137710780&partnerID=8YFLogxK

U2 - 10.1109/ICME52920.2022.9859674

DO - 10.1109/ICME52920.2022.9859674

M3 - Conference contribution

AN - SCOPUS:85137710780

SN - 978-1-6654-8564-7

T3 - Proceedings - IEEE International Conference on Multimedia and Expo

BT - Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME)

PB - IEEE Computer Society

T2 - 2022 IEEE International Conference on Multimedia and Expo, ICME 2022

Y2 - 18 July 2022 through 22 July 2022

ER -

Qiu F, Zhao S, Zhang Y, Ma R, Liu Y, Wang Z et al. Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism. In Proceedings 2022 IEEE International Conference on Multimedia and Expo (ICME). IEEE Computer Society. 2022. 22012740. (Proceedings - IEEE International Conference on Multimedia and Expo). Epub 2022 Aug 26. doi: 10.1109/ICME52920.2022.9859674

Salient Object Detection via Bilateral Feature Fusion and Score Sorting Attention Mechanism

Abstract

Publication series

Conference

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this