Comprint Image Forgery Detection and Localization using Compression Fingerprints Hannes Mareen1 Dante Vanden Bussche1 Fabrizio Guillaro2 Davide

2025-04-27 0 0 2.7MB 19 页 10玖币
侵权投诉
Comprint: Image Forgery Detection and
Localization using Compression Fingerprints
Hannes Mareen1, Dante Vanden Bussche1, Fabrizio Guillaro2, Davide
Cozzolino2, Glenn Van Wallendael1, Peter Lambert1, and Luisa Verdoliva2
1Ghent University – imec, IDLab, ELIS, Ghent, Belgium
{hannes.mareen, dante.vandenbussche, glenn.vanwallendael,
peter.lambert}@ugent.be
https://media.idlab.ugent.be
2Università degli Studi di Napoli Federico II, Naples, Italy
{fabrizio.guillaro, davide.cozzolino, verdoliv}@unina.it
https://www.grip.unina.it
Abstract. Manipulation tools that realistically edit images are widely
available, making it easy for anyone to create and spread misinformation.
In an attempt to fight fake news, forgery detection and localization meth-
ods were designed. However, existing methods struggle to accurately re-
veal manipulations found in images on the internet, i.e., in the wild. That
is because the type of forgery is typically unknown, in addition to the
tampering traces being damaged by recompression. This paper presents
Comprint, a novel forgery detection and localization method based on
the compression fingerprint or comprint. It is trained on pristine data
only, providing generalization to detect different types of manipulation.
Additionally, we propose a fusion of Comprint with the state-of-the-art
Noiseprint, which utilizes a complementary camera model fingerprint.
We carry out an extensive experimental analysis and demonstrate that
Comprint has a high level of accuracy on five evaluation datasets that
represent a wide range of manipulation types, mimicking in-the-wild cir-
cumstances. Most notably, the proposed fusion significantly outperforms
state-of-the-art reference methods. As such, Comprint and the fusion
Comprint+Noiseprint represent a promising forensics tool to analyze in-
the-wild tampered images.
Keywords: Image Forensics ·Forgery Detection ·Forgery Localization
·Deep Learning ·In-the-wild Robustness.
arXiv:2210.02227v1 [cs.CV] 5 Oct 2022
2 H. Mareen et al.
1 Introduction
Manipulating digital images is both becoming easier and more realistic using
image editing tools and AI-based software. For example, Fig. 1a shows a manip-
ulated image in which the face of Captain Jack Sparrow is replaced with that of
another person (original from Pirates of the Caribbean). This forgery was gen-
erated in a few seconds using FaceHub.live, a free AI-based tool. Although there
are many harmless applications of these powerful tools, manipulated images con-
tribute to problems such as fake news, fake evidence, and fraud. Therefore, it
is crucial to investigate forensic methods that can fact check and verify images
found on the internet, i.e., in the wild.
Forensics forgery detection and localization methods typically look at im-
perceptible traces of an image’s digital history [23]. That is because every step
in the camera acquisition and digital editing process leaves a trail of clues. For
example, the sensor of the camera introduces a unique noise pattern [16], and
compression introduces blocking artifacts [3, 4, 12, 15, 27]. Although these traces
may be invisible to the human eye, a computer can exploit them. Unfortunately,
images found in the wild are often recompressed to a lower quality, which hides
these imperceptible traces. As such, successful forgery detection is challenging.
Recent methods that rely on deep Convolutional Neural Networks (CNNs)
have shown a higher level of robustness against recompression [22,23]. However,
most of them remain intrinsically weak because they can only detect forgeries
that were seen during training. Therefore, a promising strategy for forgery detec-
tion is using one-class deep-learning methods that are only trained using pristine
data. As such, they are not limited to specific types of forgery. For example, the
Noiseprint algorithm learns to extract a fingerprint from the camera model using
deep learning [7]. Then, inconsistencies in this fingerprint reveal tampering. This
is similar to traditional methods using PRNU fingerprints, but does not require
images from the camera of the image under investigation. It is trained on im-
ages coming from many different cameras, hence being able to detect different
sources, but no specific design was made during training to take into account
the different JPEG history of such patches.
In this paper, we exploit compression artifacts and extract a different finger-
print, called Comprint. It utilizes compression artifacts, but unlike most of the
state-of-the-art works, it does not assume that real regions underwent double
compression in contrast to fake regions that underwent single compression. In-
stead, it simply assumes that the pristine regions have a different compression
history than the tampered regions. Comprint is based on one-class training with
only pristine data: images compressed using a different Quality Factor (QF) or
quantization table in the JPEG coding standard. The architecture is based on a
Siamese network that is trained to distinguish regions that were compressed dif-
ferently. As such, the deep-learning method can extract a compression fingerprint
or comprint from an image under investigation. Then, a localization algorithm
segments the distinguished regions into a heatmap. For example, Fig. 1b and
Fig. 1c show the extracted comprint and heatmap, respectively, corresponding
to the manipulated image of Fig. 1a. We demonstrate that Comprint has a high
Comprint: Image Forgery Detection and Localization 3
(a) Manipulated image using free tool FaceHub.live.
(b) Comprint: compression fingerprint.
(c) Heatmap.
Fig. 1. Realistically manipulating an image is easier than ever, such as with (a) free AI-
based tools. Therefore, we propose (b) Comprint, a forgery detection and localization
method based on compression fingerprints. Inconsistencies in the comprint indicate and
localize forgeries that can be easily visualized through a (c) heatmap.
4 H. Mareen et al.
level of robustness to in-the-wild forgeries. Additionally, since Comprint con-
tains characteristics that are complementary to Noiseprint, we propose a fusion
of these fingerprints, Comprint+Noiseprint. We demonstrate that the fusion re-
sults in a an improved performance compared to using each method individually.
2 Related Work
This section briefly discusses the main classes of forgery detection and localiza-
tion algorithms. More specifically, since this paper proposes a method based on
compression artifacts and is inspired by Noiseprint, this section mainly focuses
on work related to these aspects.
Conventional model-based methods Conventional detection methods that were
proposed before the emergence of deep learning typically rely on prior assump-
tions, which limits their applicability in the real world. They build a handcrafted
model that describes artifacts left behind by manipulations, and detect anoma-
lies in them. For example, some are based on the photo-response non-uniformity
(PRNU) noise, which is a unique noise pattern that sensors introduce, and is
used to identify camera models or specific devices. When part of an image does
not correlate with the PRNU fingerprint of the camera, it indicates forgery [16].
Although this method is powerful, it requires many images from the same cam-
era that captured the media under investigation. Therefore, it is not always
applicable in the wild. Another example is Splicebuster, which extracts expres-
sive features that capture the traces left by in-camera processing, and use those
statistics to discover potential inconsistencies caused by splicing [6].
Models based on JPEG artifacts There also exists a large variety of model-based
methods that utilize anomalies in JPEG artifacts. For example, a mismatch
in JPEG block artifacts, the JPEG grid, or the JPEG block convergence can
indicate tampering [12, 13, 15, 27]. Additionally, many methods are based on
double quantization or double JPEG compression artifacts [3, 4]. That is, these
methods assume that the authentic region is compressed twice: once before, and
once after manipulation. In contrast, the fake region is assumed to be compressed
only once (i.e., only after manipulation). Although these methods work relatively
well in the circumstances that they were designed for, they typically perform
worse in the wild. For example, images typically undergo multiple compression
steps when shared through social networks. In general, although methods based
on JPEG artifacts have shown some merit, building theoretical models that are
applicable in the wild is very challenging. That is because they are restricted by
their assumptions that do not always hold in practice.
Data-driven methods More recent methods are often data driven, in contrast to
model-driven conventional methods. As such, the challenge shifts from building
a good theoretical model towards building a suitable training dataset which en-
ables good generalisation characteristics for unseen data. In general, data-driven
摘要:

Comprint:ImageForgeryDetectionandLocalizationusingCompressionFingerprintsHannesMareen1,DanteVandenBussche1,FabrizioGuillaro2,DavideCozzolino2,GlennVanWallendael1,PeterLambert1,andLuisaVerdoliva21GhentUniversityimec,IDLab,ELIS,Ghent,Belgium{hannes.mareen,dante.vandenbussche,glenn.vanwallendael,peter...

收起<<
Comprint Image Forgery Detection and Localization using Compression Fingerprints Hannes Mareen1 Dante Vanden Bussche1 Fabrizio Guillaro2 Davide.pdf

共19页,预览4页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:19 页 大小:2.7MB 格式:PDF 时间:2025-04-27

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 19
客服
关注