Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance

Pingping ZHANG; Xu WANG; Yun ZHANG; Lin MA; Jianmin JIANG; Sam KWONG

doi:10.1007/978-3-319-77380-3_83

Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance

Pingping ZHANG, Xu WANG, Yun ZHANG, Lin MA, Jianmin JIANG, Sam KWONG

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

1 Citation (Scopus)

Abstract

In this paper, we propose a intensity guided CNN (IG-Net) model, which learns an end-to-end mapping between the intensity image and distorted depth map to the uncompressed depth map. To eliminate the undesired blocking artifacts such as discontinuities around object boundary, two branches are designed to extract the high-frequency information from intensity image and depth map, respectively. Multi-scale feature fusion and enhancement layers are introduced in the main branch to strength the edge information of the restored depth map. Performance evaluation on JPEG compression artifacts shows the effectiveness and superiority of our proposed model compared with state-of-the-art methods.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages	863-872
DOIs	https://doi.org/10.1007/978-3-319-77380-3_83
Publication status	Published - 2018
Externally published	Yes

Bibliographical note

This work was supported in part by the National Natural Science Foundation of China under Grant 61501299, 61471348, 61672443 and 61620106008, in part by the Guangdong Nature Science Foundation under Grant 2016A030310058, in part by the Shenzhen Emerging Industries of the Strategic Basic Research Project under Grants JCYJ20150525092941043, JCYJ20160226191842793, in part by the Project 2016049 supported by SZU R/D Fund, and in part by the Tencent “Rhinoceros Birds”-Scientific Research Foundation for Young Teachers of Shenzhen University.

Keywords

Compression artifacts
Convolutional neural network
JPEG compression

Access to Document

10.1007/978-3-319-77380-3_83

Cite this

@inproceedings{e33bd5a20e1143718a789f9fc59916d5,

title = "Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance",

abstract = "In this paper, we propose a intensity guided CNN (IG-Net) model, which learns an end-to-end mapping between the intensity image and distorted depth map to the uncompressed depth map. To eliminate the undesired blocking artifacts such as discontinuities around object boundary, two branches are designed to extract the high-frequency information from intensity image and depth map, respectively. Multi-scale feature fusion and enhancement layers are introduced in the main branch to strength the edge information of the restored depth map. Performance evaluation on JPEG compression artifacts shows the effectiveness and superiority of our proposed model compared with state-of-the-art methods.",

keywords = "Compression artifacts, Convolutional neural network, JPEG compression",

author = "Pingping ZHANG and Xu WANG and Yun ZHANG and Lin MA and Jianmin JIANG and Sam KWONG",

note = "This work was supported in part by the National Natural Science Foundation of China under Grant 61501299, 61471348, 61672443 and 61620106008, in part by the Guangdong Nature Science Foundation under Grant 2016A030310058, in part by the Shenzhen Emerging Industries of the Strategic Basic Research Project under Grants JCYJ20150525092941043, JCYJ20160226191842793, in part by the Project 2016049 supported by SZU R/D Fund, and in part by the Tencent “Rhinoceros Birds”-Scientific Research Foundation for Young Teachers of Shenzhen University.",

year = "2018",

doi = "10.1007/978-3-319-77380-3_83",

language = "English",

pages = "863--872",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance. / ZHANG, Pingping; WANG, Xu; ZHANG, Yun et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2018. p. 863-872.

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

TY - GEN

T1 - Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance

AU - ZHANG, Pingping

AU - WANG, Xu

AU - ZHANG, Yun

AU - MA, Lin

AU - JIANG, Jianmin

AU - KWONG, Sam

N1 - This work was supported in part by the National Natural Science Foundation of China under Grant 61501299, 61471348, 61672443 and 61620106008, in part by the Guangdong Nature Science Foundation under Grant 2016A030310058, in part by the Shenzhen Emerging Industries of the Strategic Basic Research Project under Grants JCYJ20150525092941043, JCYJ20160226191842793, in part by the Project 2016049 supported by SZU R/D Fund, and in part by the Tencent “Rhinoceros Birds”-Scientific Research Foundation for Young Teachers of Shenzhen University.

PY - 2018

Y1 - 2018

N2 - In this paper, we propose a intensity guided CNN (IG-Net) model, which learns an end-to-end mapping between the intensity image and distorted depth map to the uncompressed depth map. To eliminate the undesired blocking artifacts such as discontinuities around object boundary, two branches are designed to extract the high-frequency information from intensity image and depth map, respectively. Multi-scale feature fusion and enhancement layers are introduced in the main branch to strength the edge information of the restored depth map. Performance evaluation on JPEG compression artifacts shows the effectiveness and superiority of our proposed model compared with state-of-the-art methods.

AB - In this paper, we propose a intensity guided CNN (IG-Net) model, which learns an end-to-end mapping between the intensity image and distorted depth map to the uncompressed depth map. To eliminate the undesired blocking artifacts such as discontinuities around object boundary, two branches are designed to extract the high-frequency information from intensity image and depth map, respectively. Multi-scale feature fusion and enhancement layers are introduced in the main branch to strength the edge information of the restored depth map. Performance evaluation on JPEG compression artifacts shows the effectiveness and superiority of our proposed model compared with state-of-the-art methods.

KW - Compression artifacts

KW - Convolutional neural network

KW - JPEG compression

UR - http://www.scopus.com/inward/record.url?scp=85047467579&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-77380-3_83

DO - 10.1007/978-3-319-77380-3_83

M3 - Conference paper (refereed)

SP - 863

EP - 872

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -

Compression Artifacts Reduction for Depth Map by Deep Intensity Guidance

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this