Saliency detection via robust seed selection of foreground and background priors

Muwei Jian; Ruihong Wang; Hui Yu; Junyu Dong; Yujuan Wang; Yilong Yin; Kin-Man Lam

doi:10.1109/APSIPAASC47483.2019.9023263

Saliency detection via robust seed selection of foreground and background priors

Muwei Jian, Ruihong Wang, Hui Yu, Junyu Dong, Yujuan Wang, Yilong Yin, Kin-Man Lam

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

198 Downloads (Pure)

Abstract

Recently, saliency detection has become a research hotspot in both the computer-vision and image-processing fields. Among the diverse saliency-detection approaches, those based on the foreground and background-based model can achieve promising performance. Reliable seed selection for the foreground and background priors is a critical step for successful saliency detection. In this paper, we firstly exploit the spatial distribution of the extracted directional patches to predict the centroid of each salient object in an image. Then, we adopt the located centroids as the visual-attention center of the whole image to compute the superpixel-based center prior, which can facilitate the subsequent seed selection for the foreground and background-prior generation. Finally, the two individual foreground-based and background-based saliency maps are combined together into a plausible and authentic saliency map. Extensive experimental assessments on publicly available datasets demonstrate the effectiveness of our proposed model.

Original language	English
Title of host publication	2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	797-801
ISBN (Electronic)	978-1-7281-3248-8
ISBN (Print)	978-1-7281-3249-5
DOIs	https://doi.org/10.1109/APSIPAASC47483.2019.9023263
Publication status	Published - 5 Mar 2020
Event	2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference - Lanzhou, China Duration: 18 Nov 2019 → 21 Nov 2019 http://apsipa2019.org/

Publication series

Name	IEEE APSIPA ASC Proceedings Series
Publisher	IEEE
ISSN (Print)	2640-009X
ISSN (Electronic)	2640-0103

Conference

Conference	2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference
Abbreviated title	APSIPA 2019
Country/Territory	China
City	Lanzhou
Period	18/11/19 → 21/11/19
Internet address	http://apsipa2019.org/

Access to Document

10.1109/APSIPAASC47483.2019.9023263

Saliency Detection via Robust Seed Selection _pp
© 2020 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.
Accepted author manuscript (Post-print), 1.39 MB

Cite this

Jian, M., Wang, R., Yu, H., Dong, J., Wang, Y., Yin, Y., & Lam, K.-M. (2020). Saliency detection via robust seed selection of foreground and background priors. In 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) (pp. 797-801). (IEEE APSIPA ASC Proceedings Series). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/APSIPAASC47483.2019.9023263

@inproceedings{5a0e509fcba84ca7bca58f0e909b7769,

title = "Saliency detection via robust seed selection of foreground and background priors",

abstract = "Recently, saliency detection has become a research hotspot in both the computer-vision and image-processing fields. Among the diverse saliency-detection approaches, those based on the foreground and background-based model can achieve promising performance. Reliable seed selection for the foreground and background priors is a critical step for successful saliency detection. In this paper, we firstly exploit the spatial distribution of the extracted directional patches to predict the centroid of each salient object in an image. Then, we adopt the located centroids as the visual-attention center of the whole image to compute the superpixel-based center prior, which can facilitate the subsequent seed selection for the foreground and background-prior generation. Finally, the two individual foreground-based and background-based saliency maps are combined together into a plausible and authentic saliency map. Extensive experimental assessments on publicly available datasets demonstrate the effectiveness of our proposed model.",

author = "Muwei Jian and Ruihong Wang and Hui Yu and Junyu Dong and Yujuan Wang and Yilong Yin and Kin-Man Lam",

year = "2020",

month = mar,

day = "5",

doi = "10.1109/APSIPAASC47483.2019.9023263",

language = "English",

isbn = "978-1-7281-3249-5",

series = "IEEE APSIPA ASC Proceedings Series",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "797--801",

booktitle = "2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)",

address = "United States",

note = "2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA 2019 ; Conference date: 18-11-2019 Through 21-11-2019",

url = "http://apsipa2019.org/",

}

Jian, M, Wang, R, Yu, H, Dong, J, Wang, Y, Yin, Y & Lam, K-M 2020, Saliency detection via robust seed selection of foreground and background priors. in 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). IEEE APSIPA ASC Proceedings Series, Institute of Electrical and Electronics Engineers Inc., pp. 797-801, 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, Lanzhou, China, 18/11/19. https://doi.org/10.1109/APSIPAASC47483.2019.9023263

Saliency detection via robust seed selection of foreground and background priors. / Jian, Muwei; Wang, Ruihong; Yu, Hui et al.
2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). Institute of Electrical and Electronics Engineers Inc., 2020. p. 797-801 (IEEE APSIPA ASC Proceedings Series).

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

TY - GEN

T1 - Saliency detection via robust seed selection of foreground and background priors

AU - Jian, Muwei

AU - Wang, Ruihong

AU - Yu, Hui

AU - Dong, Junyu

AU - Wang, Yujuan

AU - Yin, Yilong

AU - Lam, Kin-Man

PY - 2020/3/5

Y1 - 2020/3/5

N2 - Recently, saliency detection has become a research hotspot in both the computer-vision and image-processing fields. Among the diverse saliency-detection approaches, those based on the foreground and background-based model can achieve promising performance. Reliable seed selection for the foreground and background priors is a critical step for successful saliency detection. In this paper, we firstly exploit the spatial distribution of the extracted directional patches to predict the centroid of each salient object in an image. Then, we adopt the located centroids as the visual-attention center of the whole image to compute the superpixel-based center prior, which can facilitate the subsequent seed selection for the foreground and background-prior generation. Finally, the two individual foreground-based and background-based saliency maps are combined together into a plausible and authentic saliency map. Extensive experimental assessments on publicly available datasets demonstrate the effectiveness of our proposed model.

AB - Recently, saliency detection has become a research hotspot in both the computer-vision and image-processing fields. Among the diverse saliency-detection approaches, those based on the foreground and background-based model can achieve promising performance. Reliable seed selection for the foreground and background priors is a critical step for successful saliency detection. In this paper, we firstly exploit the spatial distribution of the extracted directional patches to predict the centroid of each salient object in an image. Then, we adopt the located centroids as the visual-attention center of the whole image to compute the superpixel-based center prior, which can facilitate the subsequent seed selection for the foreground and background-prior generation. Finally, the two individual foreground-based and background-based saliency maps are combined together into a plausible and authentic saliency map. Extensive experimental assessments on publicly available datasets demonstrate the effectiveness of our proposed model.

U2 - 10.1109/APSIPAASC47483.2019.9023263

DO - 10.1109/APSIPAASC47483.2019.9023263

M3 - Conference contribution

SN - 978-1-7281-3249-5

T3 - IEEE APSIPA ASC Proceedings Series

SP - 797

EP - 801

BT - 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference

Y2 - 18 November 2019 through 21 November 2019

ER -

Jian M, Wang R, Yu H, Dong J, Wang Y, Yin Y et al. Saliency detection via robust seed selection of foreground and background priors. In 2019 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). Institute of Electrical and Electronics Engineers Inc. 2020. p. 797-801. (IEEE APSIPA ASC Proceedings Series). doi: 10.1109/APSIPAASC47483.2019.9023263

Saliency detection via robust seed selection of foreground and background priors

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this