基于多尺度视觉Transformer的图像篡改定位
TP391; 随着数字图像处理技术的不断发展,图像篡改不再局限于图像拼接等单一手段,而是通过图像编辑软件后处理隐藏恶意篡改痕迹,导致现有传统算法和基于深度学习的定位方法效果不佳.针对现有图像篡改算法定位精度不高的问题,文中提出了一种端到端基于多尺度视觉Transformer的图像篡改定位网络,该网络通过融合Transformer和卷积编码器来提取篡改区域与非篡改区域的特征差异,而多尺度视觉Transformer对不同尺寸图像块序列的空间信息进行建模,使得网络能适应各种形状大小的篡改区域.实验结果表明:所提出的算法在CASIA和NIST2016测试集上的F1分数分别为0.431和0.877,AU...
Saved in:
| Published in | 华南理工大学学报（自然科学版） Vol. 50; no. 6; pp. 10 - 18 |
|---|---|
| Main Authors | , , |
| Format | Journal Article |
| Language | Chinese |
| Published |
华南理工大学 计算机科学与工程学院, 广东 广州510640
01.06.2022
|
| Subjects | |
| Online Access | Get full text |
| ISSN | 1000-565X |
| DOI | 10.12141/j.issn.1000-565X.210603 |
Cover
| Abstract | TP391; 随着数字图像处理技术的不断发展,图像篡改不再局限于图像拼接等单一手段,而是通过图像编辑软件后处理隐藏恶意篡改痕迹,导致现有传统算法和基于深度学习的定位方法效果不佳.针对现有图像篡改算法定位精度不高的问题,文中提出了一种端到端基于多尺度视觉Transformer的图像篡改定位网络,该网络通过融合Transformer和卷积编码器来提取篡改区域与非篡改区域的特征差异,而多尺度视觉Transformer对不同尺寸图像块序列的空间信息进行建模,使得网络能适应各种形状大小的篡改区域.实验结果表明:所提出的算法在CASIA和NIST2016测试集上的F1分数分别为0.431和0.877,AUC值分别为0.728和0.971,相比当前的主流算法具有较为明显的性能提升;所提算法对JPEG压缩攻击具有较强的鲁棒性. |
|---|---|
| AbstractList | TP391; 随着数字图像处理技术的不断发展,图像篡改不再局限于图像拼接等单一手段,而是通过图像编辑软件后处理隐藏恶意篡改痕迹,导致现有传统算法和基于深度学习的定位方法效果不佳.针对现有图像篡改算法定位精度不高的问题,文中提出了一种端到端基于多尺度视觉Transformer的图像篡改定位网络,该网络通过融合Transformer和卷积编码器来提取篡改区域与非篡改区域的特征差异,而多尺度视觉Transformer对不同尺寸图像块序列的空间信息进行建模,使得网络能适应各种形状大小的篡改区域.实验结果表明:所提出的算法在CASIA和NIST2016测试集上的F1分数分别为0.431和0.877,AUC值分别为0.728和0.971,相比当前的主流算法具有较为明显的性能提升;所提算法对JPEG压缩攻击具有较强的鲁棒性. |
| Author | 陆璐 钟文煜 吴小坤 |
| AuthorAffiliation | 华南理工大学 计算机科学与工程学院, 广东 广州510640 |
| AuthorAffiliation_xml | – name: 华南理工大学 计算机科学与工程学院, 广东 广州510640 |
| Author_FL | ZHONG Wenyu LU Lu WU Xiaokun |
| Author_FL_xml | – sequence: 1 fullname: LU Lu – sequence: 2 fullname: ZHONG Wenyu – sequence: 3 fullname: WU Xiaokun |
| Author_xml | – sequence: 1 fullname: 陆璐 – sequence: 2 fullname: 钟文煜 – sequence: 3 fullname: 吴小坤 |
| BookMark | eNo9j7tKA0EYhaeIYIx5B1uLXf_5dy67pQRvELCJYBdm3ZmYEGdhBzG9IlioTaq1EEFMp7Yu6NNkTB7DBcXqwPngfJw10rC51YRsUAgpUka3RuHQORtSAAi44MchUhAQNUjzv1olbeeGKUAsZBKDbBLuH6t5deefS_9e-eplObtezm56hbLO5MWZLhbllX_48pf3i7en7-mHfy3nn7frZMWosdPtv2yRo92dXmc_6B7uHXS2u4GjwDAQILiUNEEhU5ZGqI3MWBZzZTKZUBYZbShQw7RKaohZiidoEHnCUi0FQtQim7-7F8oaZQf9UX5e2NrYP7XjQTaZpAiI9UvA6AdiCVtS |
| ClassificationCodes | TP391 |
| ContentType | Journal Article |
| Copyright | Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| Copyright_xml | – notice: Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| DBID | 2B. 4A8 92I 93N PSX TCJ |
| DOI | 10.12141/j.issn.1000-565X.210603 |
| DatabaseName | Wanfang Data Journals - Hong Kong WANFANG Data Centre Wanfang Data Journals 万方数据期刊 - 香港版 China Online Journals (COJ) China Online Journals (COJ) |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Sciences (General) |
| DocumentTitle_FL | Image Tampering Localization Based on Visual Multi-Scale Transformer |
| EndPage | 18 |
| ExternalDocumentID | hnlgdxxb202206002 |
| GrantInformation_xml | – fundername: (国家社会科学基金); (中山市产学研重大项目) funderid: (国家社会科学基金); (中山市产学研重大项目) |
| GroupedDBID | -03 2B. 4A8 92I 93N ALMA_UNASSIGNED_HOLDINGS CCEZO CEKLB PSX TCJ |
| ID | FETCH-LOGICAL-s1042-60657719267b4b32ef7d4d85afd79143fef101f4ea932e2db2c2f22594be76203 |
| ISSN | 1000-565X |
| IngestDate | Thu May 29 04:15:24 EDT 2025 |
| IsPeerReviewed | false |
| IsScholarly | true |
| Issue | 6 |
| Keywords | 纵横注意力 视觉Transformer 图像篡改 深度学习 |
| Language | Chinese |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-s1042-60657719267b4b32ef7d4d85afd79143fef101f4ea932e2db2c2f22594be76203 |
| PageCount | 9 |
| ParticipantIDs | wanfang_journals_hnlgdxxb202206002 |
| PublicationCentury | 2000 |
| PublicationDate | 2022-06-01 |
| PublicationDateYYYYMMDD | 2022-06-01 |
| PublicationDate_xml | – month: 06 year: 2022 text: 2022-06-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationTitle | 华南理工大学学报（自然科学版） |
| PublicationTitle_FL | Journal of South China University of Technology(Natural Science Edition) |
| PublicationYear | 2022 |
| Publisher | 华南理工大学 计算机科学与工程学院, 广东 广州510640 |
| Publisher_xml | – name: 华南理工大学 计算机科学与工程学院, 广东 广州510640 |
| SSID | ssib008679807 ssib036435713 ssib001129195 ssib051370482 ssib023167209 ssib006703757 ssib001166550 ssib000969305 ssj0000561675 ssib002039873 ssib008143609 ssib002263912 ssib020475101 |
| Score | 2.3539135 |
| Snippet | TP391; 随着数字图像处理技术的不断发展,图像篡改不再局限于图像拼接等单一手段,而是通过图像编辑软件后处理隐藏恶意篡改痕迹,导致现有传统算法和基于深度学习的定位方法... |
| SourceID | wanfang |
| SourceType | Aggregation Database |
| StartPage | 10 |
| Title | 基于多尺度视觉Transformer的图像篡改定位 |
| URI | https://d.wanfangdata.com.cn/periodical/hnlgdxxb202206002 |
| Volume | 50 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVEBS databaseName: Inspec with Full Text issn: 1000-565X databaseCode: ADMLS dateStart: 20190601 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.ebsco.com/products/research-databases/inspec-full-text omitProxy: false ssIdentifier: ssib001166550 providerName: EBSCOhost |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnR1Na9VAcKntxYtYP_CbIi4o8mqy2c9j8l4eRawXW-itvLwk7aE8wbZQelUED-rFUz2IIPamXn2gR39JY_sznNlsky1VqL0sm83Ox84kOzOb3Qkhd2SJKU0KeJGGIu9wpkTHgNvQGbIi0gMTZjzAA87zj-XcIn-4JJYmzvzydi1tbmSzw-2_nis5jVahDfSKp2T_Q7MNUmiAOugXStAwlCfSMU0FNX2axDTlWOoUW2JOTYyVJLC3BJaxpKmmsaK6qZiFQ6cV1JYqBNLcokxoYjHpiOo-3or7NA5pKqkBOsYSSS0RuOxR3fNdXAvYc7xAxSiLPLCUgRfAJhybwAVWepa7piKpjm0fGFmXao38aoCKEQ9wmEjLEuAMPSgckO1cQzXrGzQ11BhLHCAYNcGROwwFiAOTSAOxwPi7bReBnCfciROlAS1AkfvLJRBpN9u67AN-Ohnct5pJraiVlbCyrHWtGo8PGUShHQ2HEUpoSbw-dvAxwzfJ9jM06TtIc6TJ4jApTJ6yzmzlTBXmBAB3fMm3ZSLw3lnfMLm9w7WLU1u8Y8aThby2noh_tsE_y4ByELUOQ7ONc3W0tpJvbWUo50DanK5TTEnJJslU3Jt_9MQPlI1vWdDLD41_LaXwvq-zIDJa-Y4meNKeIy4V_ru5dZw1uP3SS0SHaSR1u7DBAq7Q-DTXmP-Btf0j8MuFah1bEUYqOMz6VP8lQAKEsFsnnFjctkAU2oN_iMweBByVg9GK57MunCfnXLA5E9czxzSZ2F69QKadOV-fuetyzt-7SET1Ybw3flN92qm-javx54Pdlwe7r7wJYn_nRfX-Z_X87f7Xj7_ffa--7Oz9eH2JLPbThe5cx_1RpbMe4jk8CQGHUhDUSZXxLGJFqXKeazEoc2VAhGVRgpRKXgwgrCtYnrEhK8HiG54V4DUF0WUyOXo6Kq6QmRImcZmDVNWg4LhyI8KSgRTDPMggaGBXyW039mU3Y64vH3terp2k03Vytn2Rb5DJjWebxU2IBDayW-4x-wNCYLuI |
| linkProvider | EBSCOhost |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=%E5%9F%BA%E4%BA%8E%E5%A4%9A%E5%B0%BA%E5%BA%A6%E8%A7%86%E8%A7%89Transformer%E7%9A%84%E5%9B%BE%E5%83%8F%E7%AF%A1%E6%94%B9%E5%AE%9A%E4%BD%8D&rft.jtitle=%E5%8D%8E%E5%8D%97%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6%E5%AD%A6%E6%8A%A5%EF%BC%88%E8%87%AA%E7%84%B6%E7%A7%91%E5%AD%A6%E7%89%88%EF%BC%89&rft.au=%E9%99%86%E7%92%90&rft.au=%E9%92%9F%E6%96%87%E7%85%9C&rft.au=%E5%90%B4%E5%B0%8F%E5%9D%A4&rft.date=2022-06-01&rft.pub=%E5%8D%8E%E5%8D%97%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6+%E8%AE%A1%E7%AE%97%E6%9C%BA%E7%A7%91%E5%AD%A6%E4%B8%8E%E5%B7%A5%E7%A8%8B%E5%AD%A6%E9%99%A2%2C+%E5%B9%BF%E4%B8%9C+%E5%B9%BF%E5%B7%9E510640&rft.issn=1000-565X&rft.volume=50&rft.issue=6&rft.spage=10&rft.epage=18&rft_id=info:doi/10.12141%2Fj.issn.1000-565X.210603&rft.externalDocID=hnlgdxxb202206002 |
| thumbnail_s | http://utb.summon.serialssolutions.com/2.0.0/image/custom?url=http%3A%2F%2Fwww.wanfangdata.com.cn%2Fimages%2FPeriodicalImages%2Fhnlgdxxb%2Fhnlgdxxb.jpg |