GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation
A large number of methods for forensics of image manipulation relies on detecting fingerprints in residuals or noises. Therefore, these detection methods are bound to be sensitive to noise generated by the image acquisition process, as well as any pre-processing. We show that a difference in pre-pro...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9036891/ |
| _version_ | 1830300833779023872 |
|---|---|
| author | Ludovic Darmet Kai Wang Francois Cayre |
| author_facet | Ludovic Darmet Kai Wang Francois Cayre |
| author_sort | Ludovic Darmet |
| collection | DOAJ |
| description | A large number of methods for forensics of image manipulation relies on detecting fingerprints in residuals or noises. Therefore, these detection methods are bound to be sensitive to noise generated by the image acquisition process, as well as any pre-processing. We show that a difference in pre-processing pipelines between training and testing sets induces performance losses for various classifiers. We focus on a particular pre-processing: resizing. It corresponds to a typical scenario where images may be resized (e.g., downscaled to reduce storage) prior to being manipulated. This performance loss due to pre-resizing could be troublesome but has been rarely investigated in the image forensics field. We propose a new and effective adaptation method for one state-of-the-art image manipulation detection pipeline, and we call our proposed method Gaussian mixture model Resizing Adaptation by Fine-Tuning (GRAFT). Adaptation is performed in an unsupervised fashion, i.e., without using any ground-truth label in the pre-resized testing domain, for the detection of image manipulation on very small patches. Experimental results show that the proposed GRAFT method can effectively improve the detection accuracy in this challenging scenario of unsupervised adaptation to resizing pre-processing. |
| first_indexed | 2024-12-19T08:35:04Z |
| format | Article |
| id | doaj.art-1f8d9ee40446414ca732617af5abd379 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-19T08:35:04Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-1f8d9ee40446414ca732617af5abd3792022-12-21T20:29:03ZengIEEEIEEE Access2169-35362020-01-018556195563210.1109/ACCESS.2020.29809929036891GRAFT: Unsupervised Adaptation to Resizing for Detection of Image ManipulationLudovic Darmet0https://orcid.org/0000-0001-5445-9763Kai Wang1Francois Cayre2GIPSA-Lab, CNRS, Grenoble INP, Université Grenoble Alpes, Grenoble, FranceGIPSA-Lab, CNRS, Grenoble INP, Université Grenoble Alpes, Grenoble, FranceGIPSA-Lab, CNRS, Grenoble INP, Université Grenoble Alpes, Grenoble, FranceA large number of methods for forensics of image manipulation relies on detecting fingerprints in residuals or noises. Therefore, these detection methods are bound to be sensitive to noise generated by the image acquisition process, as well as any pre-processing. We show that a difference in pre-processing pipelines between training and testing sets induces performance losses for various classifiers. We focus on a particular pre-processing: resizing. It corresponds to a typical scenario where images may be resized (e.g., downscaled to reduce storage) prior to being manipulated. This performance loss due to pre-resizing could be troublesome but has been rarely investigated in the image forensics field. We propose a new and effective adaptation method for one state-of-the-art image manipulation detection pipeline, and we call our proposed method Gaussian mixture model Resizing Adaptation by Fine-Tuning (GRAFT). Adaptation is performed in an unsupervised fashion, i.e., without using any ground-truth label in the pre-resized testing domain, for the detection of image manipulation on very small patches. Experimental results show that the proposed GRAFT method can effectively improve the detection accuracy in this challenging scenario of unsupervised adaptation to resizing pre-processing.https://ieeexplore.ieee.org/document/9036891/Image forensicsmanipulation detectionGaussian mixture modelcovariance matriceslikelihood maximizationdomain adaptation |
| spellingShingle | Ludovic Darmet Kai Wang Francois Cayre GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation IEEE Access Image forensics manipulation detection Gaussian mixture model covariance matrices likelihood maximization domain adaptation |
| title | GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation |
| title_full | GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation |
| title_fullStr | GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation |
| title_full_unstemmed | GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation |
| title_short | GRAFT: Unsupervised Adaptation to Resizing for Detection of Image Manipulation |
| title_sort | graft unsupervised adaptation to resizing for detection of image manipulation |
| topic | Image forensics manipulation detection Gaussian mixture model covariance matrices likelihood maximization domain adaptation |
| url | https://ieeexplore.ieee.org/document/9036891/ |
| work_keys_str_mv | AT ludovicdarmet graftunsupervisedadaptationtoresizingfordetectionofimagemanipulation AT kaiwang graftunsupervisedadaptationtoresizingfordetectionofimagemanipulation AT francoiscayre graftunsupervisedadaptationtoresizingfordetectionofimagemanipulation |