A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts

This paper presents methods based on convolutional neural networks (CNNs) for removing compression artifacts. We modify the Inception module for the image restoration problem and use it as a building block for constructing blind and non-blind artifact removal networks. It is known that a CNN trained...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on circuits and systems for video technology 2020-04, Vol.30 (4), p.1121-1135
Hauptverfasser: Kim, Yoonsik, Soh, Jae Woong, Park, Jaewoo, Ahn, Byeongyong, Lee, Hyun-Seung, Moon, Young-Su, Cho, Nam Ik
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 1135
container_issue 4
container_start_page 1121
container_title IEEE transactions on circuits and systems for video technology
container_volume 30
creator Kim, Yoonsik
Soh, Jae Woong
Park, Jaewoo
Ahn, Byeongyong
Lee, Hyun-Seung
Moon, Young-Su
Cho, Nam Ik
description This paper presents methods based on convolutional neural networks (CNNs) for removing compression artifacts. We modify the Inception module for the image restoration problem and use it as a building block for constructing blind and non-blind artifact removal networks. It is known that a CNN trained in a non-blind scenario (known compression quality factor) performs better than the one trained in a blind scenario (unknown factor), and our network is not an exception. However, the blind system is more practical because the compression quality factor is not always available or does not reflect the actual quality when the image is a transcoded or requantized image. Hence, in this paper, we also propose a pseudo-blind system that estimates the quality factor for a given compressed image and then applies a network that is trained with a similar quality factor. For this purpose, we propose a CNN that estimates the compression quality factor and prepare several non-blind artifact removal networks that are trained for some specific compression quality factors. We train the networks and conduct experiments on widely used compression standards, such as JPEG, MPEG-2, H.264, and HEVC. In addition, we conduct experiments for dynamically changing and transcoded videos to demonstrate the effectiveness of the quality estimation method. The experimental results show that the proposed pseudo-blind network performs better than the blind one for the various cases stated above and requires fewer computations.
doi_str_mv 10.1109/TCSVT.2019.2901919
format Article
fullrecord <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_ieee_primary_8653951</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>8653951</ieee_id><sourcerecordid>2387069562</sourcerecordid><originalsourceid>FETCH-LOGICAL-c361t-409d05cddf3ae90fde18f68f7626124b2b6aadc66cd1dfb889f1e2f0427f01823</originalsourceid><addsrcrecordid>eNo9kF1LwzAUhoMoOKd_QG8KXnfmpE2aXM7iFwwVnbsNWT6ws1tmkir-e9tNvDkf8D6Hw4PQOeAJABZX8_p1MZ8QDGJCRF9BHKARUMpzQjA97GdMIecE6DE6iXGFMZS8rEZoMc2eo-2Mz6_bZmOy2m--fNulxm9Umz3aLuxa-vbhI3M-ZOndZi_WdHqIZN71xHobbIzDOg2pcUqneIqOnGqjPfvrY_R2ezOv7_PZ091DPZ3lumCQ8hILg6k2xhXKCuyMBe4YdxUjDEi5JEumlNGMaQPGLTkXDixxuCSVw8BJMUaX-7vb4D87G5Nc-S70r0dJCl5hJigbUmSf0sHHGKyT29CsVfiRgOXgT-78ycGf_PPXQxd7qLHW_gOc0UJQKH4BWfNs_A</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2387069562</pqid></control><display><type>article</type><title>A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts</title><source>IEEE Electronic Library (IEL)</source><creator>Kim, Yoonsik ; Soh, Jae Woong ; Park, Jaewoo ; Ahn, Byeongyong ; Lee, Hyun-Seung ; Moon, Young-Su ; Cho, Nam Ik</creator><creatorcontrib>Kim, Yoonsik ; Soh, Jae Woong ; Park, Jaewoo ; Ahn, Byeongyong ; Lee, Hyun-Seung ; Moon, Young-Su ; Cho, Nam Ik</creatorcontrib><description>This paper presents methods based on convolutional neural networks (CNNs) for removing compression artifacts. We modify the Inception module for the image restoration problem and use it as a building block for constructing blind and non-blind artifact removal networks. It is known that a CNN trained in a non-blind scenario (known compression quality factor) performs better than the one trained in a blind scenario (unknown factor), and our network is not an exception. However, the blind system is more practical because the compression quality factor is not always available or does not reflect the actual quality when the image is a transcoded or requantized image. Hence, in this paper, we also propose a pseudo-blind system that estimates the quality factor for a given compressed image and then applies a network that is trained with a similar quality factor. For this purpose, we propose a CNN that estimates the compression quality factor and prepare several non-blind artifact removal networks that are trained for some specific compression quality factors. We train the networks and conduct experiments on widely used compression standards, such as JPEG, MPEG-2, H.264, and HEVC. In addition, we conduct experiments for dynamically changing and transcoded videos to demonstrate the effectiveness of the quality estimation method. The experimental results show that the proposed pseudo-blind network performs better than the blind one for the various cases stated above and requires fewer computations.</description><identifier>ISSN: 1051-8215</identifier><identifier>EISSN: 1558-2205</identifier><identifier>DOI: 10.1109/TCSVT.2019.2901919</identifier><identifier>CODEN: ITCTEM</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Artificial neural networks ; compression artifacts ; compression quality factor ; Convolutional neural network ; Decoding ; Image coding ; Image compression ; Image quality ; Image restoration ; inception ; Kernel ; Neural networks ; Q factors ; Q-factor ; Quality ; Training ; Transform coding</subject><ispartof>IEEE transactions on circuits and systems for video technology, 2020-04, Vol.30 (4), p.1121-1135</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c361t-409d05cddf3ae90fde18f68f7626124b2b6aadc66cd1dfb889f1e2f0427f01823</citedby><cites>FETCH-LOGICAL-c361t-409d05cddf3ae90fde18f68f7626124b2b6aadc66cd1dfb889f1e2f0427f01823</cites><orcidid>0000-0002-6816-4381 ; 0000-0001-5297-4649 ; 0000-0002-8367-9375 ; 0000-0001-8023-8278</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/8653951$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,776,780,792,27901,27902,54733</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/8653951$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Kim, Yoonsik</creatorcontrib><creatorcontrib>Soh, Jae Woong</creatorcontrib><creatorcontrib>Park, Jaewoo</creatorcontrib><creatorcontrib>Ahn, Byeongyong</creatorcontrib><creatorcontrib>Lee, Hyun-Seung</creatorcontrib><creatorcontrib>Moon, Young-Su</creatorcontrib><creatorcontrib>Cho, Nam Ik</creatorcontrib><title>A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts</title><title>IEEE transactions on circuits and systems for video technology</title><addtitle>TCSVT</addtitle><description>This paper presents methods based on convolutional neural networks (CNNs) for removing compression artifacts. We modify the Inception module for the image restoration problem and use it as a building block for constructing blind and non-blind artifact removal networks. It is known that a CNN trained in a non-blind scenario (known compression quality factor) performs better than the one trained in a blind scenario (unknown factor), and our network is not an exception. However, the blind system is more practical because the compression quality factor is not always available or does not reflect the actual quality when the image is a transcoded or requantized image. Hence, in this paper, we also propose a pseudo-blind system that estimates the quality factor for a given compressed image and then applies a network that is trained with a similar quality factor. For this purpose, we propose a CNN that estimates the compression quality factor and prepare several non-blind artifact removal networks that are trained for some specific compression quality factors. We train the networks and conduct experiments on widely used compression standards, such as JPEG, MPEG-2, H.264, and HEVC. In addition, we conduct experiments for dynamically changing and transcoded videos to demonstrate the effectiveness of the quality estimation method. The experimental results show that the proposed pseudo-blind network performs better than the blind one for the various cases stated above and requires fewer computations.</description><subject>Artificial neural networks</subject><subject>compression artifacts</subject><subject>compression quality factor</subject><subject>Convolutional neural network</subject><subject>Decoding</subject><subject>Image coding</subject><subject>Image compression</subject><subject>Image quality</subject><subject>Image restoration</subject><subject>inception</subject><subject>Kernel</subject><subject>Neural networks</subject><subject>Q factors</subject><subject>Q-factor</subject><subject>Quality</subject><subject>Training</subject><subject>Transform coding</subject><issn>1051-8215</issn><issn>1558-2205</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNo9kF1LwzAUhoMoOKd_QG8KXnfmpE2aXM7iFwwVnbsNWT6ws1tmkir-e9tNvDkf8D6Hw4PQOeAJABZX8_p1MZ8QDGJCRF9BHKARUMpzQjA97GdMIecE6DE6iXGFMZS8rEZoMc2eo-2Mz6_bZmOy2m--fNulxm9Umz3aLuxa-vbhI3M-ZOndZi_WdHqIZN71xHobbIzDOg2pcUqneIqOnGqjPfvrY_R2ezOv7_PZ091DPZ3lumCQ8hILg6k2xhXKCuyMBe4YdxUjDEi5JEumlNGMaQPGLTkXDixxuCSVw8BJMUaX-7vb4D87G5Nc-S70r0dJCl5hJigbUmSf0sHHGKyT29CsVfiRgOXgT-78ycGf_PPXQxd7qLHW_gOc0UJQKH4BWfNs_A</recordid><startdate>20200401</startdate><enddate>20200401</enddate><creator>Kim, Yoonsik</creator><creator>Soh, Jae Woong</creator><creator>Park, Jaewoo</creator><creator>Ahn, Byeongyong</creator><creator>Lee, Hyun-Seung</creator><creator>Moon, Young-Su</creator><creator>Cho, Nam Ik</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><orcidid>https://orcid.org/0000-0002-6816-4381</orcidid><orcidid>https://orcid.org/0000-0001-5297-4649</orcidid><orcidid>https://orcid.org/0000-0002-8367-9375</orcidid><orcidid>https://orcid.org/0000-0001-8023-8278</orcidid></search><sort><creationdate>20200401</creationdate><title>A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts</title><author>Kim, Yoonsik ; Soh, Jae Woong ; Park, Jaewoo ; Ahn, Byeongyong ; Lee, Hyun-Seung ; Moon, Young-Su ; Cho, Nam Ik</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c361t-409d05cddf3ae90fde18f68f7626124b2b6aadc66cd1dfb889f1e2f0427f01823</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Artificial neural networks</topic><topic>compression artifacts</topic><topic>compression quality factor</topic><topic>Convolutional neural network</topic><topic>Decoding</topic><topic>Image coding</topic><topic>Image compression</topic><topic>Image quality</topic><topic>Image restoration</topic><topic>inception</topic><topic>Kernel</topic><topic>Neural networks</topic><topic>Q factors</topic><topic>Q-factor</topic><topic>Quality</topic><topic>Training</topic><topic>Transform coding</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Kim, Yoonsik</creatorcontrib><creatorcontrib>Soh, Jae Woong</creatorcontrib><creatorcontrib>Park, Jaewoo</creatorcontrib><creatorcontrib>Ahn, Byeongyong</creatorcontrib><creatorcontrib>Lee, Hyun-Seung</creatorcontrib><creatorcontrib>Moon, Young-Su</creatorcontrib><creatorcontrib>Cho, Nam Ik</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE transactions on circuits and systems for video technology</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Kim, Yoonsik</au><au>Soh, Jae Woong</au><au>Park, Jaewoo</au><au>Ahn, Byeongyong</au><au>Lee, Hyun-Seung</au><au>Moon, Young-Su</au><au>Cho, Nam Ik</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts</atitle><jtitle>IEEE transactions on circuits and systems for video technology</jtitle><stitle>TCSVT</stitle><date>2020-04-01</date><risdate>2020</risdate><volume>30</volume><issue>4</issue><spage>1121</spage><epage>1135</epage><pages>1121-1135</pages><issn>1051-8215</issn><eissn>1558-2205</eissn><coden>ITCTEM</coden><abstract>This paper presents methods based on convolutional neural networks (CNNs) for removing compression artifacts. We modify the Inception module for the image restoration problem and use it as a building block for constructing blind and non-blind artifact removal networks. It is known that a CNN trained in a non-blind scenario (known compression quality factor) performs better than the one trained in a blind scenario (unknown factor), and our network is not an exception. However, the blind system is more practical because the compression quality factor is not always available or does not reflect the actual quality when the image is a transcoded or requantized image. Hence, in this paper, we also propose a pseudo-blind system that estimates the quality factor for a given compressed image and then applies a network that is trained with a similar quality factor. For this purpose, we propose a CNN that estimates the compression quality factor and prepare several non-blind artifact removal networks that are trained for some specific compression quality factors. We train the networks and conduct experiments on widely used compression standards, such as JPEG, MPEG-2, H.264, and HEVC. In addition, we conduct experiments for dynamically changing and transcoded videos to demonstrate the effectiveness of the quality estimation method. The experimental results show that the proposed pseudo-blind network performs better than the blind one for the various cases stated above and requires fewer computations.</abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/TCSVT.2019.2901919</doi><tpages>15</tpages><orcidid>https://orcid.org/0000-0002-6816-4381</orcidid><orcidid>https://orcid.org/0000-0001-5297-4649</orcidid><orcidid>https://orcid.org/0000-0002-8367-9375</orcidid><orcidid>https://orcid.org/0000-0001-8023-8278</orcidid></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1051-8215
ispartof IEEE transactions on circuits and systems for video technology, 2020-04, Vol.30 (4), p.1121-1135
issn 1051-8215
1558-2205
language eng
recordid cdi_ieee_primary_8653951
source IEEE Electronic Library (IEL)
subjects Artificial neural networks
compression artifacts
compression quality factor
Convolutional neural network
Decoding
Image coding
Image compression
Image quality
Image restoration
inception
Kernel
Neural networks
Q factors
Q-factor
Quality
Training
Transform coding
title A Pseudo-Blind Convolutional Neural Network for the Reduction of Compression Artifacts
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-31T03%3A46%3A13IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Pseudo-Blind%20Convolutional%20Neural%20Network%20for%20the%20Reduction%20of%20Compression%20Artifacts&rft.jtitle=IEEE%20transactions%20on%20circuits%20and%20systems%20for%20video%20technology&rft.au=Kim,%20Yoonsik&rft.date=2020-04-01&rft.volume=30&rft.issue=4&rft.spage=1121&rft.epage=1135&rft.pages=1121-1135&rft.issn=1051-8215&rft.eissn=1558-2205&rft.coden=ITCTEM&rft_id=info:doi/10.1109/TCSVT.2019.2901919&rft_dat=%3Cproquest_RIE%3E2387069562%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2387069562&rft_id=info:pmid/&rft_ieee_id=8653951&rfr_iscdi=true